diff options
Diffstat (limited to 'drivers/gpu/drm/amd/display/dc/clk_mgr')
15 files changed, 1569 insertions, 49 deletions
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/Makefile b/drivers/gpu/drm/amd/display/dc/clk_mgr/Makefile index 1a495759a034..d59b380e7b7f 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/Makefile +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/Makefile @@ -114,14 +114,46 @@ endif AMD_DAL_CLK_MGR_DCN21 = $(addprefix $(AMDDALPATH)/dc/clk_mgr/dcn21/,$(CLK_MGR_DCN21)) AMD_DISPLAY_FILES += $(AMD_DAL_CLK_MGR_DCN21) -endif -ifdef CONFIG_DRM_AMD_DC_DCN3_0 ############################################################################### # DCN30 ############################################################################### CLK_MGR_DCN30 = dcn30_clk_mgr.o dcn30_clk_mgr_smu_msg.o +# prevent build errors regarding soft-float vs hard-float FP ABI tags +# this code is currently unused on ppc64, as it applies to VanGogh APUs only +ifdef CONFIG_PPC64 +CFLAGS_$(AMDDALPATH)/dc/clk_mgr/dcn30/dcn30_clk_mgr.o := $(call cc-option,-mno-gnu-attribute) +endif + +# prevent build errors: +# ...: '-mgeneral-regs-only' is incompatible with the use of floating-point types +# this file is unused on arm64, just like on ppc64 +ifdef CONFIG_ARM64 +CFLAGS_REMOVE_$(AMDDALPATH)/dc/clk_mgr/dcn30/dcn30_clk_mgr.o := -mgeneral-regs-only +endif + AMD_DAL_CLK_MGR_DCN30 = $(addprefix $(AMDDALPATH)/dc/clk_mgr/dcn30/,$(CLK_MGR_DCN30)) AMD_DISPLAY_FILES += $(AMD_DAL_CLK_MGR_DCN30) +############################################################################### +# DCN301 +############################################################################### +CLK_MGR_DCN301 = vg_clk_mgr.o dcn301_smu.o + +# prevent build errors regarding soft-float vs hard-float FP ABI tags +# this code is currently unused on ppc64, as it applies to VanGogh APUs only +ifdef CONFIG_PPC64 +CFLAGS_$(AMDDALPATH)/dc/clk_mgr/dcn301/vg_clk_mgr.o := $(call cc-option,-mno-gnu-attribute) +endif + +# prevent build errors: +# ...: '-mgeneral-regs-only' is incompatible with the use of floating-point types +# this file is unused on arm64, just like on ppc64 +ifdef CONFIG_ARM64 +CFLAGS_REMOVE_$(AMDDALPATH)/dc/clk_mgr/dcn301/vg_clk_mgr.o := -mgeneral-regs-only +endif + +AMD_DAL_CLK_MGR_DCN301 = $(addprefix $(AMDDALPATH)/dc/clk_mgr/dcn301/,$(CLK_MGR_DCN301)) + +AMD_DISPLAY_FILES += $(AMD_DAL_CLK_MGR_DCN301) endif diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c index 857f156e4985..995ffbbf64e7 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c @@ -39,9 +39,8 @@ #include "dcn10/rv2_clk_mgr.h" #include "dcn20/dcn20_clk_mgr.h" #include "dcn21/rn_clk_mgr.h" -#if defined(CONFIG_DRM_AMD_DC_DCN3_0) #include "dcn30/dcn30_clk_mgr.h" -#endif +#include "dcn301/vg_clk_mgr.h" int clk_mgr_helper_get_active_display_cnt( @@ -95,7 +94,7 @@ void clk_mgr_exit_optimized_pwr_state(const struct dc *dc, struct clk_mgr *clk_m if (edp_link) { clk_mgr->psr_allow_active_cache = edp_link->psr_settings.psr_allow_active; - dc_link_set_psr_allow_active(edp_link, false, false); + dc_link_set_psr_allow_active(edp_link, false, false, false); } } @@ -105,7 +104,8 @@ void clk_mgr_optimize_pwr_state(const struct dc *dc, struct clk_mgr *clk_mgr) struct dc_link *edp_link = get_edp_link(dc); if (edp_link) - dc_link_set_psr_allow_active(edp_link, clk_mgr->psr_allow_active_cache, false); + dc_link_set_psr_allow_active(edp_link, + clk_mgr->psr_allow_active_cache, false, false); if (dc->hwss.optimize_pwr_state) dc->hwss.optimize_pwr_state(dc, dc->current_state); @@ -183,16 +183,22 @@ struct clk_mgr *dc_clk_mgr_create(struct dc_context *ctx, struct pp_smu_funcs *p break; case FAMILY_NV: -#if defined(CONFIG_DRM_AMD_DC_DCN3_0) if (ASICREV_IS_SIENNA_CICHLID_P(asic_id.hw_internal_rev)) { dcn3_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg); break; } -#endif + if (ASICREV_IS_DIMGREY_CAVEFISH_P(asic_id.hw_internal_rev)) { + dcn3_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg); + break; + } dcn20_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg); break; -#endif /* Family RV and NV*/ + case FAMILY_VGH: + if (ASICREV_IS_VANGOGH(asic_id.hw_internal_rev)) + vg_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg); + break; +#endif default: ASSERT(0); /* Unknown Asic */ break; @@ -204,14 +210,22 @@ struct clk_mgr *dc_clk_mgr_create(struct dc_context *ctx, struct pp_smu_funcs *p void dc_destroy_clk_mgr(struct clk_mgr *clk_mgr_base) { struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); -#ifdef CONFIG_DRM_AMD_DC_DCN3_0 +#ifdef CONFIG_DRM_AMD_DC_DCN switch (clk_mgr_base->ctx->asic_id.chip_family) { case FAMILY_NV: if (ASICREV_IS_SIENNA_CICHLID_P(clk_mgr_base->ctx->asic_id.hw_internal_rev)) { dcn3_clk_mgr_destroy(clk_mgr); - break; } + break; + + case FAMILY_VGH: + if (ASICREV_IS_VANGOGH(clk_mgr_base->ctx->asic_id.hw_internal_rev)) + vg_clk_mgr_destroy(clk_mgr); + break; + + default: + break; } #endif diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce110/dce110_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce110/dce110_clk_mgr.c index b1e657e137a9..78df96882d6e 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce110/dce110_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce110/dce110_clk_mgr.c @@ -195,7 +195,8 @@ void dce11_pplib_apply_display_requirements( * , then change minimum memory clock based on real-time bandwidth * limitation. */ - if (ASICREV_IS_VEGA20_P(dc->ctx->asic_id.hw_internal_rev) && (context->stream_count >= 2)) { + if ((dc->ctx->asic_id.chip_family == FAMILY_AI) && + ASICREV_IS_VEGA20_P(dc->ctx->asic_id.hw_internal_rev) && (context->stream_count >= 2)) { pp_display_cfg->min_memory_clock_khz = max(pp_display_cfg->min_memory_clock_khz, (uint32_t) div64_s64( div64_s64(dc->bw_vbios->high_yclk.value, diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce112/dce112_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce112/dce112_clk_mgr.c index 807dca8f7d7a..934e6423dc1a 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce112/dce112_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce112/dce112_clk_mgr.c @@ -100,7 +100,9 @@ int dce112_set_clock(struct clk_mgr *clk_mgr_base, int requested_clk_khz) /*VBIOS will determine DPREFCLK frequency, so we don't set it*/ dce_clk_params.target_clock_frequency = 0; dce_clk_params.clock_type = DCECLOCK_TYPE_DPREFCLK; - if (!ASICREV_IS_VEGA20_P(clk_mgr_base->ctx->asic_id.hw_internal_rev)) + + if (!((clk_mgr_base->ctx->asic_id.chip_family == FAMILY_AI) && + ASICREV_IS_VEGA20_P(clk_mgr_base->ctx->asic_id.hw_internal_rev))) dce_clk_params.flags.USE_GENLOCK_AS_SOURCE_FOR_DPREFCLK = (dce_clk_params.pll_id == CLOCK_SOURCE_COMBO_DISPLAY_PLL0); @@ -176,7 +178,8 @@ int dce112_set_dprefclk(struct clk_mgr_internal *clk_mgr) dce_clk_params.target_clock_frequency = 0; dce_clk_params.pll_id = CLOCK_SOURCE_ID_DFS; dce_clk_params.clock_type = DCECLOCK_TYPE_DPREFCLK; - if (!ASICREV_IS_VEGA20_P(clk_mgr->base.ctx->asic_id.hw_internal_rev)) + if (!((clk_mgr->base.ctx->asic_id.chip_family == FAMILY_AI) && + ASICREV_IS_VEGA20_P(clk_mgr->base.ctx->asic_id.hw_internal_rev))) dce_clk_params.flags.USE_GENLOCK_AS_SOURCE_FOR_DPREFCLK = (dce_clk_params.pll_id == CLOCK_SOURCE_COMBO_DISPLAY_PLL0); diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c index e133edc587d3..75b8240ed059 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c @@ -187,6 +187,17 @@ static void ramp_up_dispclk_with_dpp( clk_mgr->base.clks.max_supported_dppclk_khz = new_clocks->max_supported_dppclk_khz; } +static bool is_mpo_enabled(struct dc_state *context) +{ + int i; + + for (i = 0; i < context->stream_count; i++) { + if (context->stream_status[i].plane_count > 1) + return true; + } + return false; +} + static void rv1_update_clocks(struct clk_mgr *clk_mgr_base, struct dc_state *context, bool safe_to_lower) @@ -284,9 +295,22 @@ static void rv1_update_clocks(struct clk_mgr *clk_mgr_base, if (pp_smu->set_hard_min_fclk_by_freq && pp_smu->set_hard_min_dcfclk_by_freq && pp_smu->set_min_deep_sleep_dcfclk) { - pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, new_clocks->fclk_khz / 1000); - pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, new_clocks->dcfclk_khz / 1000); - pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, (new_clocks->dcfclk_deep_sleep_khz + 999) / 1000); + // Only increase clocks when display is active and MPO is enabled + if (display_count && is_mpo_enabled(context)) { + pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, + ((new_clocks->fclk_khz / 1000) * 101) / 100); + pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, + ((new_clocks->dcfclk_khz / 1000) * 101) / 100); + pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, + (new_clocks->dcfclk_deep_sleep_khz + 999) / 1000); + } else { + pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, + new_clocks->fclk_khz / 1000); + pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, + new_clocks->dcfclk_khz / 1000); + pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, + (new_clocks->dcfclk_deep_sleep_khz + 999) / 1000); + } } } } diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c index 2f8fee05547a..01b1853b7750 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c @@ -75,15 +75,8 @@ int rn_get_active_display_cnt_wa( for (i = 0; i < dc->link_count; i++) { const struct dc_link *link = dc->links[i]; - /* - * Only notify active stream or virtual stream. - * Need to notify virtual stream to work around - * headless case. HPD does not fire when system is in - * S0i2. - */ /* abusing the fact that the dig and phy are coupled to see if the phy is enabled */ - if (link->connector_signal == SIGNAL_TYPE_VIRTUAL || - link->link_enc->funcs->is_dig_enabled(link->link_enc)) + if (link->link_enc->funcs->is_dig_enabled(link->link_enc)) display_count++; } @@ -103,6 +96,31 @@ void rn_set_low_power_state(struct clk_mgr *clk_mgr_base) clk_mgr_base->clks.pwr_state = DCN_PWR_STATE_LOW_POWER; } +static void rn_update_clocks_update_dpp_dto(struct clk_mgr_internal *clk_mgr, + struct dc_state *context, int ref_dpp_clk, bool safe_to_lower) +{ + int i; + + clk_mgr->dccg->ref_dppclk = ref_dpp_clk; + + for (i = 0; i < clk_mgr->base.ctx->dc->res_pool->pipe_count; i++) { + int dpp_inst, dppclk_khz, prev_dppclk_khz; + + /* Loop index will match dpp->inst if resource exists, + * and we want to avoid dependency on dpp object + */ + dpp_inst = i; + dppclk_khz = context->res_ctx.pipe_ctx[i].plane_res.bw.dppclk_khz; + + prev_dppclk_khz = clk_mgr->dccg->pipe_dppclk_khz[i]; + + if (safe_to_lower || prev_dppclk_khz < dppclk_khz) + clk_mgr->dccg->funcs->update_dpp_dto( + clk_mgr->dccg, dpp_inst, dppclk_khz); + } +} + + void rn_update_clocks(struct clk_mgr *clk_mgr_base, struct dc_state *context, bool safe_to_lower) @@ -124,7 +142,7 @@ void rn_update_clocks(struct clk_mgr *clk_mgr_base, * if it is safe to lower, but we are already in the lower state, we don't have to do anything * also if safe to lower is false, we just go in the higher state */ - if (safe_to_lower) { + if (safe_to_lower && !dc->debug.disable_48mhz_pwrdwn) { /* check that we're not already in lower */ if (clk_mgr_base->clks.pwr_state != DCN_PWR_STATE_LOW_POWER) { @@ -158,13 +176,20 @@ void rn_update_clocks(struct clk_mgr *clk_mgr_base, // workaround: Limit dppclk to 100Mhz to avoid lower eDP panel switch to plus 4K monitor underflow. // Do not adjust dppclk if dppclk is 0 to avoid unexpected result - if (!IS_DIAG_DC(dc->ctx->dce_environment)) { - if (new_clocks->dppclk_khz < 100000 && new_clocks->dppclk_khz > 0) - new_clocks->dppclk_khz = 100000; + if (new_clocks->dppclk_khz < 100000 && new_clocks->dppclk_khz > 0) + new_clocks->dppclk_khz = 100000; + + /* + * Temporally ignore thew 0 cases for disp and dpp clks. + * We may have a new feature that requires 0 clks in the future. + */ + if (new_clocks->dppclk_khz == 0 || new_clocks->dispclk_khz == 0) { + new_clocks->dppclk_khz = clk_mgr_base->clks.dppclk_khz; + new_clocks->dispclk_khz = clk_mgr_base->clks.dispclk_khz; } - if (should_set_clock(safe_to_lower, new_clocks->dppclk_khz, clk_mgr->base.clks.dppclk_khz)) { - if (clk_mgr->base.clks.dppclk_khz > new_clocks->dppclk_khz) + if (should_set_clock(safe_to_lower, new_clocks->dppclk_khz, clk_mgr_base->clks.dppclk_khz)) { + if (clk_mgr_base->clks.dppclk_khz > new_clocks->dppclk_khz) dpp_clock_lowered = true; clk_mgr_base->clks.dppclk_khz = new_clocks->dppclk_khz; update_dppclk = true; @@ -172,22 +197,41 @@ void rn_update_clocks(struct clk_mgr *clk_mgr_base, if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz)) { clk_mgr_base->clks.dispclk_khz = new_clocks->dispclk_khz; - rn_vbios_smu_set_dispclk(clk_mgr, clk_mgr_base->clks.dispclk_khz); + clk_mgr_base->clks.actual_dispclk_khz = rn_vbios_smu_set_dispclk(clk_mgr, clk_mgr_base->clks.dispclk_khz); update_dispclk = true; } if (dpp_clock_lowered) { - // increase per DPP DTO before lowering global dppclk - dcn20_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower); - rn_vbios_smu_set_dppclk(clk_mgr, clk_mgr_base->clks.dppclk_khz); + // increase per DPP DTO before lowering global dppclk with requested dppclk + rn_update_clocks_update_dpp_dto( + clk_mgr, + context, + clk_mgr_base->clks.dppclk_khz, + safe_to_lower); + + clk_mgr_base->clks.actual_dppclk_khz = + rn_vbios_smu_set_dppclk(clk_mgr, clk_mgr_base->clks.dppclk_khz); + + //update dpp dto with actual dpp clk. + rn_update_clocks_update_dpp_dto( + clk_mgr, + context, + clk_mgr_base->clks.actual_dppclk_khz, + safe_to_lower); + } else { // increase global DPPCLK before lowering per DPP DTO if (update_dppclk || update_dispclk) - rn_vbios_smu_set_dppclk(clk_mgr, clk_mgr_base->clks.dppclk_khz); + clk_mgr_base->clks.actual_dppclk_khz = + rn_vbios_smu_set_dppclk(clk_mgr, clk_mgr_base->clks.dppclk_khz); + // always update dtos unless clock is lowered and not safe to lower - if (new_clocks->dppclk_khz >= dc->current_state->bw_ctx.bw.dcn.clk.dppclk_khz) - dcn20_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower); + rn_update_clocks_update_dpp_dto( + clk_mgr, + context, + clk_mgr_base->clks.actual_dppclk_khz, + safe_to_lower); } if (update_dispclk && @@ -198,7 +242,6 @@ void rn_update_clocks(struct clk_mgr *clk_mgr_base, } } - static int get_vco_frequency_from_reg(struct clk_mgr_internal *clk_mgr) { /* get FbMult value */ @@ -570,14 +613,14 @@ static struct clk_bw_params rn_bw_params = { }; -static struct wm_table ddr4_wm_table = { +static struct wm_table ddr4_wm_table_gs = { .entries = { { .wm_inst = WM_A, .wm_type = WM_TYPE_PSTATE_CHG, .pstate_latency_us = 11.72, - .sr_exit_time_us = 6.09, - .sr_enter_plus_exit_time_us = 7.14, + .sr_exit_time_us = 7.09, + .sr_enter_plus_exit_time_us = 8.14, .valid = true, }, { @@ -607,7 +650,7 @@ static struct wm_table ddr4_wm_table = { } }; -static struct wm_table lpddr4_wm_table = { +static struct wm_table lpddr4_wm_table_gs = { .entries = { { .wm_inst = WM_A, @@ -681,6 +724,80 @@ static struct wm_table lpddr4_wm_table_with_disabled_ppt = { } }; +static struct wm_table ddr4_wm_table_rn = { + .entries = { + { + .wm_inst = WM_A, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.72, + .sr_exit_time_us = 11.90, + .sr_enter_plus_exit_time_us = 12.80, + .valid = true, + }, + { + .wm_inst = WM_B, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.72, + .sr_exit_time_us = 13.18, + .sr_enter_plus_exit_time_us = 14.30, + .valid = true, + }, + { + .wm_inst = WM_C, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.72, + .sr_exit_time_us = 13.18, + .sr_enter_plus_exit_time_us = 14.30, + .valid = true, + }, + { + .wm_inst = WM_D, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.72, + .sr_exit_time_us = 13.18, + .sr_enter_plus_exit_time_us = 14.30, + .valid = true, + }, + } +}; + +static struct wm_table lpddr4_wm_table_rn = { + .entries = { + { + .wm_inst = WM_A, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.65333, + .sr_exit_time_us = 7.32, + .sr_enter_plus_exit_time_us = 8.38, + .valid = true, + }, + { + .wm_inst = WM_B, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.65333, + .sr_exit_time_us = 9.82, + .sr_enter_plus_exit_time_us = 11.196, + .valid = true, + }, + { + .wm_inst = WM_C, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.65333, + .sr_exit_time_us = 9.89, + .sr_enter_plus_exit_time_us = 11.24, + .valid = true, + }, + { + .wm_inst = WM_D, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.65333, + .sr_exit_time_us = 9.748, + .sr_enter_plus_exit_time_us = 11.102, + .valid = true, + }, + } +}; + static unsigned int find_dcfclk_for_voltage(struct dpm_clocks *clock_table, unsigned int voltage) { int i; @@ -762,6 +879,11 @@ void rn_clk_mgr_construct( struct dc_debug_options *debug = &ctx->dc->debug; struct dpm_clocks clock_table = { 0 }; enum pp_smu_status status = 0; + int is_green_sardine = 0; + +#if defined(CONFIG_DRM_AMD_DC_DCN) + is_green_sardine = ASICREV_IS_GREEN_SARDINE(ctx->asic_id.hw_internal_rev); +#endif clk_mgr->base.ctx = ctx; clk_mgr->base.funcs = &dcn21_funcs; @@ -802,10 +924,16 @@ void rn_clk_mgr_construct( if (clk_mgr->periodic_retraining_disabled) { rn_bw_params.wm_table = lpddr4_wm_table_with_disabled_ppt; } else { - rn_bw_params.wm_table = lpddr4_wm_table; + if (is_green_sardine) + rn_bw_params.wm_table = lpddr4_wm_table_gs; + else + rn_bw_params.wm_table = lpddr4_wm_table_rn; } } else { - rn_bw_params.wm_table = ddr4_wm_table; + if (is_green_sardine) + rn_bw_params.wm_table = ddr4_wm_table_gs; + else + rn_bw_params.wm_table = ddr4_wm_table_rn; } /* Saved clocks configured at boot for debug purposes */ rn_dump_clk_registers(&clk_mgr->base.boot_snapshot, &clk_mgr->base, &log_info); diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c index 9a374522e963..7deeec9d1c7c 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c @@ -99,7 +99,7 @@ int rn_vbios_smu_send_msg_with_param(struct clk_mgr_internal *clk_mgr, unsigned /* Trigger the message transaction by writing the message ID */ REG_WRITE(MP1_SMN_C2PMSG_67, msg_id); - result = rn_smu_wait_for_response(clk_mgr, 10, 1000); + result = rn_smu_wait_for_response(clk_mgr, 10, 200000); ASSERT(result == VBIOSSMC_Result_OK || result == VBIOSSMC_Result_UnknownCmd); @@ -136,6 +136,10 @@ int rn_vbios_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_dis } } + // pmfw always set clock more than or equal requested clock + if (!IS_DIAG_DC(dc->ctx->dce_environment)) + ASSERT(actual_dispclk_set_mhz >= requested_dispclk_khz / 1000); + return actual_dispclk_set_mhz * 1000; } @@ -194,12 +198,16 @@ void rn_vbios_smu_set_phyclk(struct clk_mgr_internal *clk_mgr, int requested_phy int rn_vbios_smu_set_dppclk(struct clk_mgr_internal *clk_mgr, int requested_dpp_khz) { int actual_dppclk_set_mhz = -1; + struct dc *dc = clk_mgr->base.ctx->dc; actual_dppclk_set_mhz = rn_vbios_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetDppclkFreq, requested_dpp_khz / 1000); + if (!IS_DIAG_DC(dc->ctx->dce_environment)) + ASSERT(actual_dppclk_set_mhz >= requested_dpp_khz / 1000); + return actual_dppclk_set_mhz * 1000; } @@ -239,5 +247,6 @@ int rn_vbios_smu_is_periodic_retraining_disabled(struct clk_mgr_internal *clk_mg return rn_vbios_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_IsPeriodicRetrainingDisabled, - 0); + 1); // if PMFW doesn't support this message, assume retraining is disabled + // so we only use most optimal watermark if we know retraining is enabled. } diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dalsmc.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dalsmc.h index 5ed03287aaaf..fa09c594fd36 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dalsmc.h +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dalsmc.h @@ -53,6 +53,7 @@ #define DALSMC_MSG_GetDcModeMaxDpmFreq 0xC #define DALSMC_MSG_SetMinDeepSleepDcefclk 0xD #define DALSMC_MSG_NumOfDisplays 0xE +#define DALSMC_MSG_SetDisplayRefreshFromMall 0xF #define DALSMC_MSG_SetExternalClientDfCstateAllow 0x10 #define DALSMC_MSG_BacoAudioD3PME 0x11 #define DALSMC_Message_Count 0x12 diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c index b0e9b0509568..5b466f440d67 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c @@ -104,7 +104,7 @@ static void dcn3_init_single_clock(struct clk_mgr_internal *clk_mgr, PPCLK_e clk } } -static void dcn3_build_wm_range_table(struct clk_mgr_internal *clk_mgr) +static noinline void dcn3_build_wm_range_table(struct clk_mgr_internal *clk_mgr) { /* defaults */ double pstate_latency_us = clk_mgr->base.ctx->dc->dml.soc.dram_clock_change_latency_us; @@ -145,6 +145,16 @@ static void dcn3_build_wm_range_table(struct clk_mgr_internal *clk_mgr) clk_mgr->base.bw_params->wm_table.nv_entries[WM_C].pmfw_breakdown.min_uclk = min_uclk_mhz; clk_mgr->base.bw_params->wm_table.nv_entries[WM_C].pmfw_breakdown.max_uclk = 0xFFFF; + /* Set D - MALL - SR enter and exit times adjusted for MALL */ +// clk_mgr->base.bw_params->wm_table.nv_entries[WM_D].valid = true; +// clk_mgr->base.bw_params->wm_table.nv_entries[WM_D].dml_input.pstate_latency_us = pstate_latency_us; +// clk_mgr->base.bw_params->wm_table.nv_entries[WM_D].dml_input.sr_exit_time_us = 2; +// clk_mgr->base.bw_params->wm_table.nv_entries[WM_D].dml_input.sr_enter_plus_exit_time_us = 4; +// clk_mgr->base.bw_params->wm_table.nv_entries[WM_D].pmfw_breakdown.wm_type = WATERMARKS_MALL; +// clk_mgr->base.bw_params->wm_table.nv_entries[WM_D].pmfw_breakdown.min_dcfclk = 0; +// clk_mgr->base.bw_params->wm_table.nv_entries[WM_D].pmfw_breakdown.max_dcfclk = 0xFFFF; +// clk_mgr->base.bw_params->wm_table.nv_entries[WM_D].pmfw_breakdown.min_uclk = min_uclk_mhz; +// clk_mgr->base.bw_params->wm_table.nv_entries[WM_D].pmfw_breakdown.max_uclk = 0xFFFF; } void dcn3_init_clocks(struct clk_mgr *clk_mgr_base) @@ -201,7 +211,9 @@ void dcn3_init_clocks(struct clk_mgr *clk_mgr_base) clk_mgr_base->funcs->get_memclk_states_from_smu(clk_mgr_base); /* WM range table */ + DC_FP_START(); dcn3_build_wm_range_table(clk_mgr); + DC_FP_END(); } static int dcn30_get_vco_frequency_from_reg(struct clk_mgr_internal *clk_mgr) @@ -487,7 +499,7 @@ static void dcn3_init_clocks_fpga(struct clk_mgr *clk_mgr) /* TODO: Implement the functions and remove the ifndef guard */ } -static struct clk_mgr_funcs dcn3_fpga_funcs = { +struct clk_mgr_funcs dcn3_fpga_funcs = { .get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz, .update_clocks = dcn2_update_clocks_fpga, .init_clocks = dcn3_init_clocks_fpga, diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr_smu_msg.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr_smu_msg.c index 7ee3ec5a8af8..8ecc708bcd9e 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr_smu_msg.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr_smu_msg.c @@ -297,6 +297,15 @@ void dcn30_smu_set_num_of_displays(struct clk_mgr_internal *clk_mgr, uint32_t nu DALSMC_MSG_NumOfDisplays, num_displays, NULL); } +void dcn30_smu_set_display_refresh_from_mall(struct clk_mgr_internal *clk_mgr, bool enable, uint8_t cache_timer_delay, uint8_t cache_timer_scale) +{ + /* bits 8:7 for cache timer scale, bits 6:1 for cache timer delay, bit 0 = 1 for enable, = 0 for disable */ + uint32_t param = (cache_timer_scale << 7) | (cache_timer_delay << 1) | (enable ? 1 : 0); + + dcn30_smu_send_msg_with_param(clk_mgr, + DALSMC_MSG_SetDisplayRefreshFromMall, param, NULL); +} + void dcn30_smu_set_external_client_df_cstate_allow(struct clk_mgr_internal *clk_mgr, bool enable) { smu_print("SMU Set external client df cstate allow: enable = %d\n", enable); diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr_smu_msg.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr_smu_msg.h index 236f20ec90d4..dd2640a3ce5d 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr_smu_msg.h +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr_smu_msg.h @@ -70,6 +70,7 @@ typedef enum { typedef enum { WATERMARKS_CLOCK_RANGE = 0, WATERMARKS_DUMMY_PSTATE, + WATERMARKS_MALL, WATERMARKS_COUNT, } WATERMARKS_FLAGS_e; @@ -102,6 +103,7 @@ unsigned int dcn30_smu_get_dpm_freq_by_index(struct clk_mgr_internal *clk_mgr, P unsigned int dcn30_smu_get_dc_mode_max_dpm_freq(struct clk_mgr_internal *clk_mgr, PPCLK_e clk); void dcn30_smu_set_min_deep_sleep_dcef_clk(struct clk_mgr_internal *clk_mgr, uint32_t freq_mhz); void dcn30_smu_set_num_of_displays(struct clk_mgr_internal *clk_mgr, uint32_t num_displays); +void dcn30_smu_set_display_refresh_from_mall(struct clk_mgr_internal *clk_mgr, bool enable, uint8_t cache_timer_delay, uint8_t cache_timer_scale); void dcn30_smu_set_external_client_df_cstate_allow(struct clk_mgr_internal *clk_mgr, bool enable); void dcn30_smu_set_pme_workaround(struct clk_mgr_internal *clk_mgr); diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/dcn301_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/dcn301_smu.c new file mode 100644 index 000000000000..cfa8e02cf103 --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/dcn301_smu.c @@ -0,0 +1,241 @@ +/* + * Copyright 2020 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#include "core_types.h" +#include "clk_mgr_internal.h" +#include "reg_helper.h" +#include <linux/delay.h> + +#include "dcn301_smu.h" + +#include "vangogh_ip_offset.h" + +#include "mp/mp_11_5_0_offset.h" +#include "mp/mp_11_5_0_sh_mask.h" + +#define REG(reg_name) \ + (MP0_BASE.instance[0].segment[mm ## reg_name ## _BASE_IDX] + mm ## reg_name) + +#define FN(reg_name, field) \ + FD(reg_name##__##field) + +#define VBIOSSMC_MSG_GetSmuVersion 0x2 +#define VBIOSSMC_MSG_SetDispclkFreq 0x4 +#define VBIOSSMC_MSG_SetDprefclkFreq 0x5 +#define VBIOSSMC_MSG_SetDppclkFreq 0x6 +#define VBIOSSMC_MSG_SetHardMinDcfclkByFreq 0x7 +#define VBIOSSMC_MSG_SetMinDeepSleepDcfclk 0x8 +//#define VBIOSSMC_MSG_SetPhyclkVoltageByFreq 0xA +#define VBIOSSMC_MSG_GetFclkFrequency 0xA +//#define VBIOSSMC_MSG_SetDisplayCount 0xC +//#define VBIOSSMC_MSG_EnableTmdp48MHzRefclkPwrDown 0xD +#define VBIOSSMC_MSG_UpdatePmeRestore 0xD +#define VBIOSSMC_MSG_SetVbiosDramAddrHigh 0xE //Used for WM table txfr +#define VBIOSSMC_MSG_SetVbiosDramAddrLow 0xF +#define VBIOSSMC_MSG_TransferTableSmu2Dram 0x10 +#define VBIOSSMC_MSG_TransferTableDram2Smu 0x11 +#define VBIOSSMC_MSG_SetDisplayIdleOptimizations 0x12 + +#define VBIOSSMC_Status_BUSY 0x0 +#define VBIOSSMC_Result_OK 0x1 +#define VBIOSSMC_Result_Failed 0xFF +#define VBIOSSMC_Result_UnknownCmd 0xFE +#define VBIOSSMC_Result_CmdRejectedPrereq 0xFD +#define VBIOSSMC_Result_CmdRejectedBusy 0xFC + +/* + * Function to be used instead of REG_WAIT macro because the wait ends when + * the register is NOT EQUAL to zero, and because the translation in msg_if.h + * won't work with REG_WAIT. + */ +static uint32_t dcn301_smu_wait_for_response(struct clk_mgr_internal *clk_mgr, unsigned int delay_us, unsigned int max_retries) +{ + uint32_t res_val = VBIOSSMC_Status_BUSY; + + do { + res_val = REG_READ(MP1_SMN_C2PMSG_91); + if (res_val != VBIOSSMC_Status_BUSY) + break; + + if (delay_us >= 1000) + msleep(delay_us/1000); + else if (delay_us > 0) + udelay(delay_us); + } while (max_retries--); + + return res_val; +} + +int dcn301_smu_send_msg_with_param( + struct clk_mgr_internal *clk_mgr, + unsigned int msg_id, unsigned int param) +{ + uint32_t result; + + /* First clear response register */ + REG_WRITE(MP1_SMN_C2PMSG_91, VBIOSSMC_Status_BUSY); + + /* Set the parameter register for the SMU message, unit is Mhz */ + REG_WRITE(MP1_SMN_C2PMSG_83, param); + + /* Trigger the message transaction by writing the message ID */ + REG_WRITE(MP1_SMN_C2PMSG_67, msg_id); + + result = dcn301_smu_wait_for_response(clk_mgr, 10, 1000); + + ASSERT(result == VBIOSSMC_Result_OK); + + /* Actual dispclk set is returned in the parameter register */ + return REG_READ(MP1_SMN_C2PMSG_83); +} + +int dcn301_smu_get_smu_version(struct clk_mgr_internal *clk_mgr) +{ + return dcn301_smu_send_msg_with_param( + clk_mgr, + VBIOSSMC_MSG_GetSmuVersion, + 0); +} + + +int dcn301_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_dispclk_khz) +{ + int actual_dispclk_set_mhz = -1; + + /* Unit of SMU msg parameter is Mhz */ + actual_dispclk_set_mhz = dcn301_smu_send_msg_with_param( + clk_mgr, + VBIOSSMC_MSG_SetDispclkFreq, + requested_dispclk_khz / 1000); + + return actual_dispclk_set_mhz * 1000; +} + +int dcn301_smu_set_dprefclk(struct clk_mgr_internal *clk_mgr) +{ + int actual_dprefclk_set_mhz = -1; + + actual_dprefclk_set_mhz = dcn301_smu_send_msg_with_param( + clk_mgr, + VBIOSSMC_MSG_SetDprefclkFreq, + clk_mgr->base.dprefclk_khz / 1000); + + /* TODO: add code for programing DP DTO, currently this is down by command table */ + + return actual_dprefclk_set_mhz * 1000; +} + +int dcn301_smu_set_hard_min_dcfclk(struct clk_mgr_internal *clk_mgr, int requested_dcfclk_khz) +{ + int actual_dcfclk_set_mhz = -1; + + actual_dcfclk_set_mhz = dcn301_smu_send_msg_with_param( + clk_mgr, + VBIOSSMC_MSG_SetHardMinDcfclkByFreq, + requested_dcfclk_khz / 1000); + + return actual_dcfclk_set_mhz * 1000; +} + +int dcn301_smu_set_min_deep_sleep_dcfclk(struct clk_mgr_internal *clk_mgr, int requested_min_ds_dcfclk_khz) +{ + int actual_min_ds_dcfclk_mhz = -1; + + actual_min_ds_dcfclk_mhz = dcn301_smu_send_msg_with_param( + clk_mgr, + VBIOSSMC_MSG_SetMinDeepSleepDcfclk, + requested_min_ds_dcfclk_khz / 1000); + + return actual_min_ds_dcfclk_mhz * 1000; +} + +int dcn301_smu_set_dppclk(struct clk_mgr_internal *clk_mgr, int requested_dpp_khz) +{ + int actual_dppclk_set_mhz = -1; + + actual_dppclk_set_mhz = dcn301_smu_send_msg_with_param( + clk_mgr, + VBIOSSMC_MSG_SetDppclkFreq, + requested_dpp_khz / 1000); + + return actual_dppclk_set_mhz * 1000; +} + +void dcn301_smu_set_display_idle_optimization(struct clk_mgr_internal *clk_mgr, uint32_t idle_info) +{ + //TODO: Work with smu team to define optimization options. + + dcn301_smu_send_msg_with_param( + clk_mgr, + VBIOSSMC_MSG_SetDisplayIdleOptimizations, + idle_info); +} + +void dcn301_smu_enable_phy_refclk_pwrdwn(struct clk_mgr_internal *clk_mgr, bool enable) +{ + union display_idle_optimization_u idle_info = { 0 }; + + if (enable) { + idle_info.idle_info.df_request_disabled = 1; + idle_info.idle_info.phy_ref_clk_off = 1; + } + + dcn301_smu_send_msg_with_param( + clk_mgr, + VBIOSSMC_MSG_SetDisplayIdleOptimizations, + idle_info.data); +} + +void dcn301_smu_enable_pme_wa(struct clk_mgr_internal *clk_mgr) +{ + dcn301_smu_send_msg_with_param( + clk_mgr, + VBIOSSMC_MSG_UpdatePmeRestore, + 0); +} + +void dcn301_smu_set_dram_addr_high(struct clk_mgr_internal *clk_mgr, uint32_t addr_high) +{ + dcn301_smu_send_msg_with_param(clk_mgr, + VBIOSSMC_MSG_SetVbiosDramAddrHigh, addr_high); +} + +void dcn301_smu_set_dram_addr_low(struct clk_mgr_internal *clk_mgr, uint32_t addr_low) +{ + dcn301_smu_send_msg_with_param(clk_mgr, + VBIOSSMC_MSG_SetVbiosDramAddrLow, addr_low); +} + +void dcn301_smu_transfer_dpm_table_smu_2_dram(struct clk_mgr_internal *clk_mgr) +{ + dcn301_smu_send_msg_with_param(clk_mgr, + VBIOSSMC_MSG_TransferTableSmu2Dram, TABLE_DPMCLOCKS); +} + +void dcn301_smu_transfer_wm_table_dram_2_smu(struct clk_mgr_internal *clk_mgr) +{ + dcn301_smu_send_msg_with_param(clk_mgr, + VBIOSSMC_MSG_TransferTableDram2Smu, TABLE_WATERMARKS); +} diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/dcn301_smu.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/dcn301_smu.h new file mode 100644 index 000000000000..b640df85a17f --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/dcn301_smu.h @@ -0,0 +1,164 @@ +/* + * Copyright 2020 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#ifndef DAL_DC_301_SMU_H_ +#define DAL_DC_301_SMU_H_ + +#define SMU13_DRIVER_IF_VERSION 2 + +typedef struct { + uint32_t fclk; + uint32_t memclk; + uint32_t voltage; +} df_pstate_t; + +typedef struct { + uint32_t vclk; + uint32_t dclk; +} vcn_clk_t; + +typedef enum { + DSPCLK_DCFCLK = 0, + DSPCLK_DISPCLK, + DSPCLK_PIXCLK, + DSPCLK_PHYCLK, + DSPCLK_COUNT, +} DSPCLK_e; + +typedef struct { + uint16_t Freq; // in MHz + uint16_t Vid; // min voltage in SVI2 VID +} DisplayClockTable_t; + +typedef struct { + uint16_t MinClock; // This is either DCFCLK or SOCCLK (in MHz) + uint16_t MaxClock; // This is either DCFCLK or SOCCLK (in MHz) + uint16_t MinMclk; + uint16_t MaxMclk; + + uint8_t WmSetting; + uint8_t WmType; // Used for normal pstate change or memory retraining + uint8_t Padding[2]; +} WatermarkRowGeneric_t; + + +#define NUM_WM_RANGES 4 + +typedef enum { + WM_SOCCLK = 0, + WM_DCFCLK, + WM_COUNT, +} WM_CLOCK_e; + +typedef struct { + // Watermarks + WatermarkRowGeneric_t WatermarkRow[WM_COUNT][NUM_WM_RANGES]; + + uint32_t MmHubPadding[7]; // SMU internal use +} Watermarks_t; + + +#define TABLE_WATERMARKS 1 +#define TABLE_DPMCLOCKS 4 // Called by Driver + + +#define VG_NUM_DCFCLK_DPM_LEVELS 7 +#define VG_NUM_DISPCLK_DPM_LEVELS 7 +#define VG_NUM_DPPCLK_DPM_LEVELS 7 +#define VG_NUM_SOCCLK_DPM_LEVELS 7 +#define VG_NUM_ISPICLK_DPM_LEVELS 7 +#define VG_NUM_ISPXCLK_DPM_LEVELS 7 +#define VG_NUM_VCN_DPM_LEVELS 5 +#define VG_NUM_FCLK_DPM_LEVELS 4 +#define VG_NUM_SOC_VOLTAGE_LEVELS 8 + +// copy from vgh/vangogh/pmfw_driver_if.h +struct vg_dpm_clocks { + uint32_t DcfClocks[VG_NUM_DCFCLK_DPM_LEVELS]; + uint32_t DispClocks[VG_NUM_DISPCLK_DPM_LEVELS]; + uint32_t DppClocks[VG_NUM_DPPCLK_DPM_LEVELS]; + uint32_t SocClocks[VG_NUM_SOCCLK_DPM_LEVELS]; + uint32_t IspiClocks[VG_NUM_ISPICLK_DPM_LEVELS]; + uint32_t IspxClocks[VG_NUM_ISPXCLK_DPM_LEVELS]; + vcn_clk_t VcnClocks[VG_NUM_VCN_DPM_LEVELS]; + + uint32_t SocVoltage[VG_NUM_SOC_VOLTAGE_LEVELS]; + + df_pstate_t DfPstateTable[VG_NUM_FCLK_DPM_LEVELS]; + + uint32_t MinGfxClk; + uint32_t MaxGfxClk; + + uint8_t NumDfPstatesEnabled; + uint8_t NumDcfclkLevelsEnabled; + uint8_t NumDispClkLevelsEnabled; //applies to both dispclk and dppclk + uint8_t NumSocClkLevelsEnabled; + + uint8_t IspClkLevelsEnabled; //applies to both ispiclk and ispxclk + uint8_t VcnClkLevelsEnabled; //applies to both vclk/dclk + uint8_t spare[2]; +}; + +struct smu_dpm_clks { + struct vg_dpm_clocks *dpm_clks; + union large_integer mc_address; +}; + +struct watermarks { + // Watermarks + WatermarkRowGeneric_t WatermarkRow[WM_COUNT][NUM_WM_RANGES]; + + uint32_t MmHubPadding[7]; // SMU internal use +}; + + +struct display_idle_optimization { + unsigned int df_request_disabled : 1; + unsigned int phy_ref_clk_off : 1; + unsigned int s0i2_rdy : 1; + unsigned int reserved : 29; +}; + +union display_idle_optimization_u { + struct display_idle_optimization idle_info; + uint32_t data; +}; + + +int dcn301_smu_get_smu_version(struct clk_mgr_internal *clk_mgr); +int dcn301_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_dispclk_khz); +int dcn301_smu_set_dprefclk(struct clk_mgr_internal *clk_mgr); +int dcn301_smu_set_hard_min_dcfclk(struct clk_mgr_internal *clk_mgr, int requested_dcfclk_khz); +int dcn301_smu_set_min_deep_sleep_dcfclk(struct clk_mgr_internal *clk_mgr, int requested_min_ds_dcfclk_khz); +int dcn301_smu_set_dppclk(struct clk_mgr_internal *clk_mgr, int requested_dpp_khz); +void dcn301_smu_set_display_idle_optimization(struct clk_mgr_internal *clk_mgr, uint32_t idle_info); +void dcn301_smu_enable_phy_refclk_pwrdwn(struct clk_mgr_internal *clk_mgr, bool enable); +void dcn301_smu_enable_pme_wa(struct clk_mgr_internal *clk_mgr); +void dcn301_smu_set_dram_addr_high(struct clk_mgr_internal *clk_mgr, uint32_t addr_high); +void dcn301_smu_set_dram_addr_low(struct clk_mgr_internal *clk_mgr, uint32_t addr_low); +void dcn301_smu_transfer_dpm_table_smu_2_dram(struct clk_mgr_internal *clk_mgr); +void dcn301_smu_transfer_wm_table_dram_2_smu(struct clk_mgr_internal *clk_mgr); + +#endif /* DAL_DC_301_SMU_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c new file mode 100644 index 000000000000..991b9c5beaa3 --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c @@ -0,0 +1,827 @@ +/* + * Copyright 2020 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#include "dccg.h" +#include "clk_mgr_internal.h" + +// For dce12_get_dp_ref_freq_khz +#include "dce100/dce_clk_mgr.h" + +// For dcn20_update_clocks_update_dpp_dto +#include "dcn20/dcn20_clk_mgr.h" + + + +#include "vg_clk_mgr.h" +#include "reg_helper.h" +#include "core_types.h" +#include "dm_helpers.h" + +#include "atomfirmware.h" +#include "vangogh_ip_offset.h" +#include "clk/clk_11_5_0_offset.h" +#include "clk/clk_11_5_0_sh_mask.h" + +/* Constants */ + +#define LPDDR_MEM_RETRAIN_LATENCY 4.977 /* Number obtained from LPDDR4 Training Counter Requirement doc */ + +/* Macros */ + +#define REG(reg_name) \ + (CLK_BASE.instance[0].segment[mm ## reg_name ## _BASE_IDX] + mm ## reg_name) + +/* TODO: evaluate how to lower or disable all dcn clocks in screen off case */ +int vg_get_active_display_cnt_wa( + struct dc *dc, + struct dc_state *context) +{ + int i, display_count; + bool tmds_present = false; + + display_count = 0; + for (i = 0; i < context->stream_count; i++) { + const struct dc_stream_state *stream = context->streams[i]; + + if (stream->signal == SIGNAL_TYPE_HDMI_TYPE_A || + stream->signal == SIGNAL_TYPE_DVI_SINGLE_LINK || + stream->signal == SIGNAL_TYPE_DVI_DUAL_LINK) + tmds_present = true; + } + + for (i = 0; i < dc->link_count; i++) { + const struct dc_link *link = dc->links[i]; + + /* abusing the fact that the dig and phy are coupled to see if the phy is enabled */ + if (link->link_enc->funcs->is_dig_enabled(link->link_enc)) + display_count++; + } + + /* WA for hang on HDMI after display off back back on*/ + if (display_count == 0 && tmds_present) + display_count = 1; + + return display_count; +} + +void vg_update_clocks(struct clk_mgr *clk_mgr_base, + struct dc_state *context, + bool safe_to_lower) +{ + struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); + struct dc_clocks *new_clocks = &context->bw_ctx.bw.dcn.clk; + struct dc *dc = clk_mgr_base->ctx->dc; + int display_count; + bool update_dppclk = false; + bool update_dispclk = false; + bool dpp_clock_lowered = false; + + if (dc->work_arounds.skip_clock_update) + return; + + /* + * if it is safe to lower, but we are already in the lower state, we don't have to do anything + * also if safe to lower is false, we just go in the higher state + */ + if (safe_to_lower) { + /* check that we're not already in lower */ + if (clk_mgr_base->clks.pwr_state != DCN_PWR_STATE_LOW_POWER) { + + display_count = vg_get_active_display_cnt_wa(dc, context); + /* if we can go lower, go lower */ + if (display_count == 0) { + union display_idle_optimization_u idle_info = { 0 }; + + idle_info.idle_info.df_request_disabled = 1; + idle_info.idle_info.phy_ref_clk_off = 1; + + dcn301_smu_set_display_idle_optimization(clk_mgr, idle_info.data); + /* update power state */ + clk_mgr_base->clks.pwr_state = DCN_PWR_STATE_LOW_POWER; + } + } + } else { + /* check that we're not already in D0 */ + if (clk_mgr_base->clks.pwr_state != DCN_PWR_STATE_MISSION_MODE) { + union display_idle_optimization_u idle_info = { 0 }; + + dcn301_smu_set_display_idle_optimization(clk_mgr, idle_info.data); + /* update power state */ + clk_mgr_base->clks.pwr_state = DCN_PWR_STATE_MISSION_MODE; + } + } + + if (should_set_clock(safe_to_lower, new_clocks->dcfclk_khz, clk_mgr_base->clks.dcfclk_khz)) { + clk_mgr_base->clks.dcfclk_khz = new_clocks->dcfclk_khz; + dcn301_smu_set_hard_min_dcfclk(clk_mgr, clk_mgr_base->clks.dcfclk_khz); + } + + if (should_set_clock(safe_to_lower, + new_clocks->dcfclk_deep_sleep_khz, clk_mgr_base->clks.dcfclk_deep_sleep_khz)) { + clk_mgr_base->clks.dcfclk_deep_sleep_khz = new_clocks->dcfclk_deep_sleep_khz; + dcn301_smu_set_min_deep_sleep_dcfclk(clk_mgr, clk_mgr_base->clks.dcfclk_deep_sleep_khz); + } + + // workaround: Limit dppclk to 100Mhz to avoid lower eDP panel switch to plus 4K monitor underflow. + if (!IS_DIAG_DC(dc->ctx->dce_environment)) { + if (new_clocks->dppclk_khz < 100000) + new_clocks->dppclk_khz = 100000; + } + + if (should_set_clock(safe_to_lower, new_clocks->dppclk_khz, clk_mgr->base.clks.dppclk_khz)) { + if (clk_mgr->base.clks.dppclk_khz > new_clocks->dppclk_khz) + dpp_clock_lowered = true; + clk_mgr_base->clks.dppclk_khz = new_clocks->dppclk_khz; + update_dppclk = true; + } + + if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz)) { + clk_mgr_base->clks.dispclk_khz = new_clocks->dispclk_khz; + dcn301_smu_set_dispclk(clk_mgr, clk_mgr_base->clks.dispclk_khz); + + update_dispclk = true; + } + + if (dpp_clock_lowered) { + // increase per DPP DTO before lowering global dppclk + dcn20_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower); + dcn301_smu_set_dppclk(clk_mgr, clk_mgr_base->clks.dppclk_khz); + } else { + // increase global DPPCLK before lowering per DPP DTO + if (update_dppclk || update_dispclk) + dcn301_smu_set_dppclk(clk_mgr, clk_mgr_base->clks.dppclk_khz); + // always update dtos unless clock is lowered and not safe to lower + if (new_clocks->dppclk_khz >= dc->current_state->bw_ctx.bw.dcn.clk.dppclk_khz) + dcn20_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower); + } +} + + +static int get_vco_frequency_from_reg(struct clk_mgr_internal *clk_mgr) +{ + /* get FbMult value */ + struct fixed31_32 pll_req; + unsigned int fbmult_frac_val = 0; + unsigned int fbmult_int_val = 0; + + + /* + * Register value of fbmult is in 8.16 format, we are converting to 31.32 + * to leverage the fix point operations available in driver + */ + + REG_GET(CLK1_0_CLK1_CLK_PLL_REQ, FbMult_frac, &fbmult_frac_val); /* 16 bit fractional part*/ + REG_GET(CLK1_0_CLK1_CLK_PLL_REQ, FbMult_int, &fbmult_int_val); /* 8 bit integer part */ + + pll_req = dc_fixpt_from_int(fbmult_int_val); + + /* + * since fractional part is only 16 bit in register definition but is 32 bit + * in our fix point definiton, need to shift left by 16 to obtain correct value + */ + pll_req.value |= fbmult_frac_val << 16; + + /* multiply by REFCLK period */ + pll_req = dc_fixpt_mul_int(pll_req, clk_mgr->dfs_ref_freq_khz); + + /* integer part is now VCO frequency in kHz */ + return dc_fixpt_floor(pll_req); +} + +static void vg_dump_clk_registers_internal(struct dcn301_clk_internal *internal, struct clk_mgr *clk_mgr_base) +{ + struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); + + internal->CLK1_CLK3_CURRENT_CNT = REG_READ(CLK1_0_CLK1_CLK3_CURRENT_CNT); + internal->CLK1_CLK3_BYPASS_CNTL = REG_READ(CLK1_0_CLK1_CLK3_BYPASS_CNTL); + + internal->CLK1_CLK3_DS_CNTL = REG_READ(CLK1_0_CLK1_CLK3_DS_CNTL); //dcf deep sleep divider + internal->CLK1_CLK3_ALLOW_DS = REG_READ(CLK1_0_CLK1_CLK3_ALLOW_DS); + + internal->CLK1_CLK1_CURRENT_CNT = REG_READ(CLK1_0_CLK1_CLK1_CURRENT_CNT); + internal->CLK1_CLK1_BYPASS_CNTL = REG_READ(CLK1_0_CLK1_CLK1_BYPASS_CNTL); + + internal->CLK1_CLK2_CURRENT_CNT = REG_READ(CLK1_0_CLK1_CLK2_CURRENT_CNT); + internal->CLK1_CLK2_BYPASS_CNTL = REG_READ(CLK1_0_CLK1_CLK2_BYPASS_CNTL); + + internal->CLK1_CLK0_CURRENT_CNT = REG_READ(CLK1_0_CLK1_CLK0_CURRENT_CNT); + internal->CLK1_CLK0_BYPASS_CNTL = REG_READ(CLK1_0_CLK1_CLK0_BYPASS_CNTL); +} + +/* This function collect raw clk register values */ +static void vg_dump_clk_registers(struct clk_state_registers_and_bypass *regs_and_bypass, + struct clk_mgr *clk_mgr_base, struct clk_log_info *log_info) +{ + struct dcn301_clk_internal internal = {0}; + char *bypass_clks[5] = {"0x0 DFS", "0x1 REFCLK", "0x2 ERROR", "0x3 400 FCH", "0x4 600 FCH"}; + unsigned int chars_printed = 0; + unsigned int remaining_buffer = log_info->bufSize; + + vg_dump_clk_registers_internal(&internal, clk_mgr_base); + + regs_and_bypass->dcfclk = internal.CLK1_CLK3_CURRENT_CNT / 10; + regs_and_bypass->dcf_deep_sleep_divider = internal.CLK1_CLK3_DS_CNTL / 10; + regs_and_bypass->dcf_deep_sleep_allow = internal.CLK1_CLK3_ALLOW_DS; + regs_and_bypass->dprefclk = internal.CLK1_CLK2_CURRENT_CNT / 10; + regs_and_bypass->dispclk = internal.CLK1_CLK0_CURRENT_CNT / 10; + regs_and_bypass->dppclk = internal.CLK1_CLK1_CURRENT_CNT / 10; + + regs_and_bypass->dppclk_bypass = internal.CLK1_CLK1_BYPASS_CNTL & 0x0007; + if (regs_and_bypass->dppclk_bypass < 0 || regs_and_bypass->dppclk_bypass > 4) + regs_and_bypass->dppclk_bypass = 0; + regs_and_bypass->dcfclk_bypass = internal.CLK1_CLK3_BYPASS_CNTL & 0x0007; + if (regs_and_bypass->dcfclk_bypass < 0 || regs_and_bypass->dcfclk_bypass > 4) + regs_and_bypass->dcfclk_bypass = 0; + regs_and_bypass->dispclk_bypass = internal.CLK1_CLK0_BYPASS_CNTL & 0x0007; + if (regs_and_bypass->dispclk_bypass < 0 || regs_and_bypass->dispclk_bypass > 4) + regs_and_bypass->dispclk_bypass = 0; + regs_and_bypass->dprefclk_bypass = internal.CLK1_CLK2_BYPASS_CNTL & 0x0007; + if (regs_and_bypass->dprefclk_bypass < 0 || regs_and_bypass->dprefclk_bypass > 4) + regs_and_bypass->dprefclk_bypass = 0; + + if (log_info->enabled) { + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "clk_type,clk_value,deepsleep_cntl,deepsleep_allow,bypass\n"); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "dcfclk,%d,%d,%d,%s\n", + regs_and_bypass->dcfclk, + regs_and_bypass->dcf_deep_sleep_divider, + regs_and_bypass->dcf_deep_sleep_allow, + bypass_clks[(int) regs_and_bypass->dcfclk_bypass]); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "dprefclk,%d,N/A,N/A,%s\n", + regs_and_bypass->dprefclk, + bypass_clks[(int) regs_and_bypass->dprefclk_bypass]); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "dispclk,%d,N/A,N/A,%s\n", + regs_and_bypass->dispclk, + bypass_clks[(int) regs_and_bypass->dispclk_bypass]); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + //split + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "SPLIT\n"); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + // REGISTER VALUES + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "reg_name,value,clk_type\n"); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "CLK1_CLK3_CURRENT_CNT,%d,dcfclk\n", + internal.CLK1_CLK3_CURRENT_CNT); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "CLK1_CLK3_DS_CNTL,%d,dcf_deep_sleep_divider\n", + internal.CLK1_CLK3_DS_CNTL); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "CLK1_CLK3_ALLOW_DS,%d,dcf_deep_sleep_allow\n", + internal.CLK1_CLK3_ALLOW_DS); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "CLK1_CLK2_CURRENT_CNT,%d,dprefclk\n", + internal.CLK1_CLK2_CURRENT_CNT); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "CLK1_CLK0_CURRENT_CNT,%d,dispclk\n", + internal.CLK1_CLK0_CURRENT_CNT); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "CLK1_CLK1_CURRENT_CNT,%d,dppclk\n", + internal.CLK1_CLK1_CURRENT_CNT); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "CLK1_CLK3_BYPASS_CNTL,%d,dcfclk_bypass\n", + internal.CLK1_CLK3_BYPASS_CNTL); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "CLK1_CLK2_BYPASS_CNTL,%d,dprefclk_bypass\n", + internal.CLK1_CLK2_BYPASS_CNTL); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "CLK1_CLK0_BYPASS_CNTL,%d,dispclk_bypass\n", + internal.CLK1_CLK0_BYPASS_CNTL); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + + chars_printed = snprintf_count(log_info->pBuf, remaining_buffer, "CLK1_CLK1_BYPASS_CNTL,%d,dppclk_bypass\n", + internal.CLK1_CLK1_BYPASS_CNTL); + remaining_buffer -= chars_printed; + *log_info->sum_chars_printed += chars_printed; + log_info->pBuf += chars_printed; + } +} + +/* This function produce translated logical clk state values*/ +void vg_get_clk_states(struct clk_mgr *clk_mgr_base, struct clk_states *s) +{ + + struct clk_state_registers_and_bypass sb = { 0 }; + struct clk_log_info log_info = { 0 }; + + vg_dump_clk_registers(&sb, clk_mgr_base, &log_info); + + s->dprefclk_khz = sb.dprefclk * 1000; +} + +void vg_enable_pme_wa(struct clk_mgr *clk_mgr_base) +{ + struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); + + dcn301_smu_enable_pme_wa(clk_mgr); +} + +void vg_init_clocks(struct clk_mgr *clk_mgr) +{ + memset(&(clk_mgr->clks), 0, sizeof(struct dc_clocks)); + // Assumption is that boot state always supports pstate + clk_mgr->clks.p_state_change_support = true; + clk_mgr->clks.prev_p_state_change_support = true; + clk_mgr->clks.pwr_state = DCN_PWR_STATE_UNKNOWN; +} + +static void vg_build_watermark_ranges(struct clk_bw_params *bw_params, struct watermarks *table) +{ + int i, num_valid_sets; + + num_valid_sets = 0; + + for (i = 0; i < WM_SET_COUNT; i++) { + /* skip empty entries, the smu array has no holes*/ + if (!bw_params->wm_table.entries[i].valid) + continue; + + table->WatermarkRow[WM_DCFCLK][num_valid_sets].WmSetting = bw_params->wm_table.entries[i].wm_inst; + table->WatermarkRow[WM_DCFCLK][num_valid_sets].WmType = bw_params->wm_table.entries[i].wm_type; + /* We will not select WM based on fclk, so leave it as unconstrained */ + table->WatermarkRow[WM_DCFCLK][num_valid_sets].MinClock = 0; + table->WatermarkRow[WM_DCFCLK][num_valid_sets].MaxClock = 0xFFFF; + + if (table->WatermarkRow[WM_DCFCLK][num_valid_sets].WmType == WM_TYPE_PSTATE_CHG) { + if (i == 0) + table->WatermarkRow[WM_DCFCLK][num_valid_sets].MinMclk = 0; + else { + /* add 1 to make it non-overlapping with next lvl */ + table->WatermarkRow[WM_DCFCLK][num_valid_sets].MinMclk = + bw_params->clk_table.entries[i - 1].dcfclk_mhz + 1; + } + table->WatermarkRow[WM_DCFCLK][num_valid_sets].MaxMclk = + bw_params->clk_table.entries[i].dcfclk_mhz; + + } else { + /* unconstrained for memory retraining */ + table->WatermarkRow[WM_DCFCLK][num_valid_sets].MinClock = 0; + table->WatermarkRow[WM_DCFCLK][num_valid_sets].MaxClock = 0xFFFF; + + /* Modify previous watermark range to cover up to max */ + table->WatermarkRow[WM_DCFCLK][num_valid_sets - 1].MaxClock = 0xFFFF; + } + num_valid_sets++; + } + + ASSERT(num_valid_sets != 0); /* Must have at least one set of valid watermarks */ + + /* modify the min and max to make sure we cover the whole range*/ + table->WatermarkRow[WM_DCFCLK][0].MinMclk = 0; + table->WatermarkRow[WM_DCFCLK][0].MinClock = 0; + table->WatermarkRow[WM_DCFCLK][num_valid_sets - 1].MaxMclk = 0xFFFF; + table->WatermarkRow[WM_DCFCLK][num_valid_sets - 1].MaxClock = 0xFFFF; + + /* This is for writeback only, does not matter currently as no writeback support*/ + table->WatermarkRow[WM_SOCCLK][0].WmSetting = WM_A; + table->WatermarkRow[WM_SOCCLK][0].MinClock = 0; + table->WatermarkRow[WM_SOCCLK][0].MaxClock = 0xFFFF; + table->WatermarkRow[WM_SOCCLK][0].MinMclk = 0; + table->WatermarkRow[WM_SOCCLK][0].MaxMclk = 0xFFFF; +} + + +void vg_notify_wm_ranges(struct clk_mgr *clk_mgr_base) +{ + struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); + struct watermarks *table = clk_mgr_base->smu_wm_set.wm_set; + + if (!clk_mgr->smu_ver) + return; + + if (!table || clk_mgr_base->smu_wm_set.mc_address.quad_part == 0) + return; + + memset(table, 0, sizeof(*table)); + + vg_build_watermark_ranges(clk_mgr_base->bw_params, table); + + dcn301_smu_set_dram_addr_high(clk_mgr, + clk_mgr_base->smu_wm_set.mc_address.high_part); + dcn301_smu_set_dram_addr_low(clk_mgr, + clk_mgr_base->smu_wm_set.mc_address.low_part); + dcn301_smu_transfer_wm_table_dram_2_smu(clk_mgr); +} + +static bool vg_are_clock_states_equal(struct dc_clocks *a, + struct dc_clocks *b) +{ + if (a->dispclk_khz != b->dispclk_khz) + return false; + else if (a->dppclk_khz != b->dppclk_khz) + return false; + else if (a->dcfclk_khz != b->dcfclk_khz) + return false; + else if (a->dcfclk_deep_sleep_khz != b->dcfclk_deep_sleep_khz) + return false; + + return true; +} + + +static struct clk_mgr_funcs vg_funcs = { + .get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz, + .update_clocks = vg_update_clocks, + .init_clocks = vg_init_clocks, + .enable_pme_wa = vg_enable_pme_wa, + .are_clock_states_equal = vg_are_clock_states_equal, + .notify_wm_ranges = vg_notify_wm_ranges +}; + +static struct clk_bw_params vg_bw_params = { + .vram_type = Ddr4MemType, + .num_channels = 1, + .clk_table = { + .entries = { + { + .voltage = 0, + .dcfclk_mhz = 400, + .fclk_mhz = 400, + .memclk_mhz = 800, + .socclk_mhz = 0, + }, + { + .voltage = 0, + .dcfclk_mhz = 483, + .fclk_mhz = 800, + .memclk_mhz = 1600, + .socclk_mhz = 0, + }, + { + .voltage = 0, + .dcfclk_mhz = 602, + .fclk_mhz = 1067, + .memclk_mhz = 1067, + .socclk_mhz = 0, + }, + { + .voltage = 0, + .dcfclk_mhz = 738, + .fclk_mhz = 1333, + .memclk_mhz = 1600, + .socclk_mhz = 0, + }, + }, + + .num_entries = 4, + }, + +}; + +static struct wm_table ddr4_wm_table = { + .entries = { + { + .wm_inst = WM_A, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.72, + .sr_exit_time_us = 6.09, + .sr_enter_plus_exit_time_us = 7.14, + .valid = true, + }, + { + .wm_inst = WM_B, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.72, + .sr_exit_time_us = 10.12, + .sr_enter_plus_exit_time_us = 11.48, + .valid = true, + }, + { + .wm_inst = WM_C, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.72, + .sr_exit_time_us = 10.12, + .sr_enter_plus_exit_time_us = 11.48, + .valid = true, + }, + { + .wm_inst = WM_D, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.72, + .sr_exit_time_us = 10.12, + .sr_enter_plus_exit_time_us = 11.48, + .valid = true, + }, + } +}; + +static struct wm_table lpddr5_wm_table = { + .entries = { + { + .wm_inst = WM_A, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.65333, + .sr_exit_time_us = 5.32, + .sr_enter_plus_exit_time_us = 6.38, + .valid = true, + }, + { + .wm_inst = WM_B, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.65333, + .sr_exit_time_us = 9.82, + .sr_enter_plus_exit_time_us = 11.196, + .valid = true, + }, + { + .wm_inst = WM_C, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.65333, + .sr_exit_time_us = 9.89, + .sr_enter_plus_exit_time_us = 11.24, + .valid = true, + }, + { + .wm_inst = WM_D, + .wm_type = WM_TYPE_PSTATE_CHG, + .pstate_latency_us = 11.65333, + .sr_exit_time_us = 9.748, + .sr_enter_plus_exit_time_us = 11.102, + .valid = true, + }, + } +}; + + +static unsigned int find_dcfclk_for_voltage(const struct vg_dpm_clocks *clock_table, + unsigned int voltage) +{ + int i; + + for (i = 0; i < VG_NUM_SOC_VOLTAGE_LEVELS; i++) { + if (clock_table->SocVoltage[i] == voltage) + return clock_table->DcfClocks[i]; + } + + ASSERT(0); + return 0; +} + +void vg_clk_mgr_helper_populate_bw_params( + struct clk_mgr_internal *clk_mgr, + struct integrated_info *bios_info, + const struct vg_dpm_clocks *clock_table) +{ + int i, j; + struct clk_bw_params *bw_params = clk_mgr->base.bw_params; + + j = -1; + + ASSERT(VG_NUM_FCLK_DPM_LEVELS <= MAX_NUM_DPM_LVL); + + /* Find lowest DPM, FCLK is filled in reverse order*/ + + for (i = VG_NUM_FCLK_DPM_LEVELS - 1; i >= 0; i--) { + if (clock_table->DfPstateTable[i].fclk != 0) { + j = i; + break; + } + } + + if (j == -1) { + /* clock table is all 0s, just use our own hardcode */ + ASSERT(0); + return; + } + + bw_params->clk_table.num_entries = j + 1; + + for (i = 0; i < bw_params->clk_table.num_entries; i++, j--) { + bw_params->clk_table.entries[i].fclk_mhz = clock_table->DfPstateTable[j].fclk; + bw_params->clk_table.entries[i].memclk_mhz = clock_table->DfPstateTable[j].memclk; + bw_params->clk_table.entries[i].voltage = clock_table->DfPstateTable[j].voltage; + bw_params->clk_table.entries[i].dcfclk_mhz = find_dcfclk_for_voltage(clock_table, clock_table->DfPstateTable[j].voltage); + } + + bw_params->vram_type = bios_info->memory_type; + bw_params->num_channels = bios_info->ma_channel_number; + + for (i = 0; i < WM_SET_COUNT; i++) { + bw_params->wm_table.entries[i].wm_inst = i; + + if (i >= bw_params->clk_table.num_entries) { + bw_params->wm_table.entries[i].valid = false; + continue; + } + + bw_params->wm_table.entries[i].wm_type = WM_TYPE_PSTATE_CHG; + bw_params->wm_table.entries[i].valid = true; + } + + if (bw_params->vram_type == LpDdr4MemType) { + /* + * WM set D will be re-purposed for memory retraining + */ + bw_params->wm_table.entries[WM_D].pstate_latency_us = LPDDR_MEM_RETRAIN_LATENCY; + bw_params->wm_table.entries[WM_D].wm_inst = WM_D; + bw_params->wm_table.entries[WM_D].wm_type = WM_TYPE_RETRAINING; + bw_params->wm_table.entries[WM_D].valid = true; + } + +} + +/* Temporary Place holder until we can get them from fuse */ +static struct vg_dpm_clocks dummy_clocks = { + .DcfClocks = { 201, 403, 403, 403, 403, 403, 403 }, + .SocClocks = { 400, 600, 600, 600, 600, 600, 600 }, + .SocVoltage = { 2800, 2860, 2860, 2860, 2860, 2860, 2860, 2860 }, + .DfPstateTable = { + { .fclk = 400, .memclk = 400, .voltage = 2800 }, + { .fclk = 400, .memclk = 400, .voltage = 2800 }, + { .fclk = 400, .memclk = 400, .voltage = 2800 }, + { .fclk = 400, .memclk = 400, .voltage = 2800 } + } +}; + +static struct watermarks dummy_wms = { 0 }; + +void vg_get_dpm_table_from_smu(struct clk_mgr_internal *clk_mgr, + struct smu_dpm_clks *smu_dpm_clks) +{ + struct vg_dpm_clocks *table = smu_dpm_clks->dpm_clks; + + if (!clk_mgr->smu_ver) + return; + + if (!table || smu_dpm_clks->mc_address.quad_part == 0) + return; + + memset(table, 0, sizeof(*table)); + + dcn301_smu_set_dram_addr_high(clk_mgr, + smu_dpm_clks->mc_address.high_part); + dcn301_smu_set_dram_addr_low(clk_mgr, + smu_dpm_clks->mc_address.low_part); + dcn301_smu_transfer_dpm_table_smu_2_dram(clk_mgr); +} + +void vg_clk_mgr_construct( + struct dc_context *ctx, + struct clk_mgr_internal *clk_mgr, + struct pp_smu_funcs *pp_smu, + struct dccg *dccg) +{ + struct smu_dpm_clks smu_dpm_clks = { 0 }; + + clk_mgr->base.ctx = ctx; + clk_mgr->base.funcs = &vg_funcs; + + clk_mgr->pp_smu = pp_smu; + + clk_mgr->dccg = dccg; + clk_mgr->dfs_bypass_disp_clk = 0; + + clk_mgr->dprefclk_ss_percentage = 0; + clk_mgr->dprefclk_ss_divider = 1000; + clk_mgr->ss_on_dprefclk = false; + clk_mgr->dfs_ref_freq_khz = 48000; + + clk_mgr->base.smu_wm_set.wm_set = (struct watermarks *)dm_helpers_allocate_gpu_mem( + clk_mgr->base.ctx, + DC_MEM_ALLOC_TYPE_FRAME_BUFFER, + sizeof(struct watermarks), + &clk_mgr->base.smu_wm_set.mc_address.quad_part); + + if (clk_mgr->base.smu_wm_set.wm_set == 0) { + clk_mgr->base.smu_wm_set.wm_set = &dummy_wms; + clk_mgr->base.smu_wm_set.mc_address.quad_part = 0; + } + ASSERT(clk_mgr->base.smu_wm_set.wm_set); + + smu_dpm_clks.dpm_clks = (struct vg_dpm_clocks *)dm_helpers_allocate_gpu_mem( + clk_mgr->base.ctx, + DC_MEM_ALLOC_TYPE_FRAME_BUFFER, + sizeof(struct vg_dpm_clocks), + &smu_dpm_clks.mc_address.quad_part); + + if (smu_dpm_clks.dpm_clks == NULL) { + smu_dpm_clks.dpm_clks = &dummy_clocks; + smu_dpm_clks.mc_address.quad_part = 0; + } + + ASSERT(smu_dpm_clks.dpm_clks); + + if (IS_FPGA_MAXIMUS_DC(ctx->dce_environment)) { + vg_funcs.update_clocks = dcn2_update_clocks_fpga; + clk_mgr->base.dentist_vco_freq_khz = 3600000; + } else { + struct clk_log_info log_info = {0}; + + clk_mgr->smu_ver = dcn301_smu_get_smu_version(clk_mgr); + + if (clk_mgr->smu_ver) + clk_mgr->smu_present = true; + + /* TODO: Check we get what we expect during bringup */ + clk_mgr->base.dentist_vco_freq_khz = get_vco_frequency_from_reg(clk_mgr); + + /* in case we don't get a value from the register, use default */ + if (clk_mgr->base.dentist_vco_freq_khz == 0) + clk_mgr->base.dentist_vco_freq_khz = 3600000; + + if (ctx->dc_bios->integrated_info->memory_type == LpDdr5MemType) { + vg_bw_params.wm_table = lpddr5_wm_table; + } else { + vg_bw_params.wm_table = ddr4_wm_table; + } + /* Saved clocks configured at boot for debug purposes */ + vg_dump_clk_registers(&clk_mgr->base.boot_snapshot, &clk_mgr->base, &log_info); + } + + clk_mgr->base.dprefclk_khz = 600000; + dce_clock_read_ss_info(clk_mgr); + + clk_mgr->base.bw_params = &vg_bw_params; + + vg_get_dpm_table_from_smu(clk_mgr, &smu_dpm_clks); + if (ctx->dc_bios && ctx->dc_bios->integrated_info) { + vg_clk_mgr_helper_populate_bw_params( + clk_mgr, + ctx->dc_bios->integrated_info, + smu_dpm_clks.dpm_clks); + } + + if (smu_dpm_clks.dpm_clks && smu_dpm_clks.mc_address.quad_part != 0) + dm_helpers_free_gpu_mem(clk_mgr->base.ctx, DC_MEM_ALLOC_TYPE_FRAME_BUFFER, + smu_dpm_clks.dpm_clks); +/* + if (!IS_FPGA_MAXIMUS_DC(ctx->dce_environment) && clk_mgr->smu_ver) { + enable powerfeatures when displaycount goes to 0 + dcn301_smu_enable_phy_refclk_pwrdwn(clk_mgr, !debug->disable_48mhz_pwrdwn); + } +*/ +} + +void vg_clk_mgr_destroy(struct clk_mgr_internal *clk_mgr) +{ + if (clk_mgr->base.smu_wm_set.wm_set && clk_mgr->base.smu_wm_set.mc_address.quad_part != 0) + dm_helpers_free_gpu_mem(clk_mgr->base.ctx, DC_MEM_ALLOC_TYPE_FRAME_BUFFER, + clk_mgr->base.smu_wm_set.wm_set); +} diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.h new file mode 100644 index 000000000000..b5115b3123a1 --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.h @@ -0,0 +1,53 @@ +/* + * Copyright 2020 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#ifndef __VG_CLK_MGR_H__ +#define __VG_CLK_MGR_H__ + +int vg_get_active_display_cnt_wa( + struct dc *dc, + struct dc_state *context); + +void vg_enable_pme_wa(struct clk_mgr *clk_mgr_base); + +void vg_clk_mgr_construct(struct dc_context *ctx, + struct clk_mgr_internal *clk_mgr, + struct pp_smu_funcs *pp_smu, + struct dccg *dccg); + +void vg_clk_mgr_destroy(struct clk_mgr_internal *clk_mgr); + +#include "dcn301_smu.h" +void vg_notify_wm_ranges(struct clk_mgr *clk_mgr_base); + +void vg_get_dpm_table_from_smu(struct clk_mgr_internal *clk_mgr, + struct smu_dpm_clks *smu_dpm_clks); + +void vg_clk_mgr_helper_populate_bw_params( + struct clk_mgr_internal *clk_mgr, + struct integrated_info *bios_info, + const struct vg_dpm_clocks *clock_table); + +#endif //__VG_CLK_MGR_H__ |