diff options
Diffstat (limited to 'drivers/gpu/drm/amd/display/dc/dml')
19 files changed, 376 insertions, 107 deletions
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c index c26da3bb2892..d3ba65efe1d2 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c @@ -26,12 +26,12 @@ #include "resource.h" #include "clk_mgr.h" -#include "dc_link_dp.h" #include "dchubbub.h" #include "dcn20/dcn20_resource.h" #include "dcn21/dcn21_resource.h" #include "clk_mgr/dcn21/rn_clk_mgr.h" +#include "link.h" #include "dcn20_fpu.h" #define DC_LOGGER_INIT(logger) @@ -938,7 +938,7 @@ static bool is_dtbclk_required(struct dc *dc, struct dc_state *context) for (i = 0; i < dc->res_pool->pipe_count; i++) { if (!context->res_ctx.pipe_ctx[i].stream) continue; - if (is_dp_128b_132b_signal(&context->res_ctx.pipe_ctx[i])) + if (link_is_dp_128b_132b_signal(&context->res_ctx.pipe_ctx[i])) return true; } return false; @@ -949,7 +949,6 @@ static enum dcn_zstate_support_state decide_zstate_support(struct dc *dc, struc int plane_count; int i; unsigned int optimized_min_dst_y_next_start_us; - bool allow_z8 = context->bw_ctx.dml.vba.StutterPeriod > 1000.0; plane_count = 0; optimized_min_dst_y_next_start_us = 0; @@ -974,6 +973,8 @@ static enum dcn_zstate_support_state decide_zstate_support(struct dc *dc, struc else if (context->stream_count == 1 && context->streams[0]->signal == SIGNAL_TYPE_EDP) { struct dc_link *link = context->streams[0]->sink->link; struct dc_stream_status *stream_status = &context->stream_status[0]; + bool allow_z8 = context->bw_ctx.dml.vba.StutterPeriod > 1000.0; + bool is_pwrseq0 = link->link_index == 0; if (dc_extended_blank_supported(dc)) { for (i = 0; i < dc->res_pool->pipe_count; i++) { @@ -986,23 +987,55 @@ static enum dcn_zstate_support_state decide_zstate_support(struct dc *dc, struc } } } - /* zstate only supported on PWRSEQ0 and when there's <2 planes*/ - if (link->link_index != 0 || stream_status->plane_count > 1) + + /* Don't support multi-plane configurations */ + if (stream_status->plane_count > 1) return DCN_ZSTATE_SUPPORT_DISALLOW; - if (context->bw_ctx.dml.vba.StutterPeriod > 5000.0 || optimized_min_dst_y_next_start_us > 5000) + if (is_pwrseq0 && (context->bw_ctx.dml.vba.StutterPeriod > 5000.0 || optimized_min_dst_y_next_start_us > 5000)) return DCN_ZSTATE_SUPPORT_ALLOW; - else if (link->psr_settings.psr_version == DC_PSR_VERSION_1 && !link->panel_config.psr.disable_psr) + else if (is_pwrseq0 && link->psr_settings.psr_version == DC_PSR_VERSION_1 && !link->panel_config.psr.disable_psr) return allow_z8 ? DCN_ZSTATE_SUPPORT_ALLOW_Z8_Z10_ONLY : DCN_ZSTATE_SUPPORT_ALLOW_Z10_ONLY; else return allow_z8 ? DCN_ZSTATE_SUPPORT_ALLOW_Z8_ONLY : DCN_ZSTATE_SUPPORT_DISALLOW; - } else if (allow_z8) { - return DCN_ZSTATE_SUPPORT_ALLOW_Z8_ONLY; } else { return DCN_ZSTATE_SUPPORT_DISALLOW; } } +static void dcn20_adjust_freesync_v_startup( + const struct dc_crtc_timing *dc_crtc_timing, int *vstartup_start) +{ + struct dc_crtc_timing patched_crtc_timing; + uint32_t asic_blank_end = 0; + uint32_t asic_blank_start = 0; + uint32_t newVstartup = 0; + + patched_crtc_timing = *dc_crtc_timing; + + if (patched_crtc_timing.flags.INTERLACE == 1) { + if (patched_crtc_timing.v_front_porch < 2) + patched_crtc_timing.v_front_porch = 2; + } else { + if (patched_crtc_timing.v_front_porch < 1) + patched_crtc_timing.v_front_porch = 1; + } + + /* blank_start = frame end - front porch */ + asic_blank_start = patched_crtc_timing.v_total - + patched_crtc_timing.v_front_porch; + + /* blank_end = blank_start - active */ + asic_blank_end = asic_blank_start - + patched_crtc_timing.v_border_bottom - + patched_crtc_timing.v_addressable - + patched_crtc_timing.v_border_top; + + newVstartup = asic_blank_end + (patched_crtc_timing.v_total - asic_blank_start); + + *vstartup_start = ((newVstartup > *vstartup_start) ? newVstartup : *vstartup_start); +} + void dcn20_calculate_dlg_params( struct dc *dc, struct dc_state *context, display_e2e_pipe_params_st *pipes, @@ -1062,6 +1095,11 @@ void dcn20_calculate_dlg_params( context->res_ctx.pipe_ctx[i].plane_res.bw.dppclk_khz = pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000; context->res_ctx.pipe_ctx[i].pipe_dlg_param = pipes[pipe_idx].pipe.dest; + if (context->res_ctx.pipe_ctx[i].stream->adaptive_sync_infopacket.valid) + dcn20_adjust_freesync_v_startup( + &context->res_ctx.pipe_ctx[i].stream->timing, + &context->res_ctx.pipe_ctx[i].pipe_dlg_param.vstartup_start); + pipe_idx++; } /*save a original dppclock copy*/ @@ -1302,7 +1340,7 @@ int dcn20_populate_dml_pipes_from_context( case SIGNAL_TYPE_DISPLAY_PORT_MST: case SIGNAL_TYPE_DISPLAY_PORT: pipes[pipe_cnt].dout.output_type = dm_dp; - if (is_dp_128b_132b_signal(&res_ctx->pipe_ctx[i])) + if (link_is_dp_128b_132b_signal(&res_ctx->pipe_ctx[i])) pipes[pipe_cnt].dout.output_type = dm_dp2p0; break; case SIGNAL_TYPE_EDP: diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c index d3b5b6fedf04..6266b0788387 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c @@ -3897,14 +3897,14 @@ void dml20_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine = mode_lib->vba.PixelClock[k] / 2 * (1 + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading / 100.0); - locals->ODMCombineEnablePerState[i][k] = false; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_disabled; mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine; if (mode_lib->vba.ODMCapability) { if (locals->PlaneRequiredDISPCLKWithoutODMCombine > mode_lib->vba.MaxDispclkRoundedDownToDFSGranularity) { - locals->ODMCombineEnablePerState[i][k] = true; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_2to1; mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine; } else if (locals->HActive[k] > DCN20_MAX_420_IMAGE_WIDTH && locals->OutputFormat[k] == dm_420) { - locals->ODMCombineEnablePerState[i][k] = true; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_2to1; mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine; } } @@ -3957,7 +3957,7 @@ void dml20_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l locals->RequiredDISPCLK[i][j] = 0.0; locals->DISPCLK_DPPCLK_Support[i][j] = true; for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - locals->ODMCombineEnablePerState[i][k] = false; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_disabled; if (locals->SwathWidthYSingleDPP[k] <= locals->MaximumSwathWidth[k]) { locals->NoOfDPP[i][j][k] = 1; locals->RequiredDPPCLK[i][j][k] = locals->MinDPPCLKUsingSingleDPP[k] diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c index edd098c7eb92..989d83ee3842 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20v2.c @@ -4008,17 +4008,17 @@ void dml20v2_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine = mode_lib->vba.PixelClock[k] / 2 * (1 + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading / 100.0); - locals->ODMCombineEnablePerState[i][k] = false; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_disabled; mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine; if (mode_lib->vba.ODMCapability) { if (locals->PlaneRequiredDISPCLKWithoutODMCombine > MaxMaxDispclkRoundedDown) { - locals->ODMCombineEnablePerState[i][k] = true; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_2to1; mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine; } else if (locals->DSCEnabled[k] && (locals->HActive[k] > DCN20_MAX_DSC_IMAGE_WIDTH)) { - locals->ODMCombineEnablePerState[i][k] = true; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_2to1; mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine; } else if (locals->HActive[k] > DCN20_MAX_420_IMAGE_WIDTH && locals->OutputFormat[k] == dm_420) { - locals->ODMCombineEnablePerState[i][k] = true; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_2to1; mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine; } } @@ -4071,7 +4071,7 @@ void dml20v2_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode locals->RequiredDISPCLK[i][j] = 0.0; locals->DISPCLK_DPPCLK_Support[i][j] = true; for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - locals->ODMCombineEnablePerState[i][k] = false; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_disabled; if (locals->SwathWidthYSingleDPP[k] <= locals->MaximumSwathWidth[k]) { locals->NoOfDPP[i][j][k] = 1; locals->RequiredDPPCLK[i][j][k] = locals->MinDPPCLKUsingSingleDPP[k] diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c index 1d84ae50311d..b7c2844d0cbe 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c @@ -4102,17 +4102,17 @@ void dml21_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine = mode_lib->vba.PixelClock[k] / 2 * (1 + mode_lib->vba.DISPCLKDPPCLKDSCCLKDownSpreading / 100.0); - locals->ODMCombineEnablePerState[i][k] = false; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_disabled; mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithoutODMCombine; if (mode_lib->vba.ODMCapability) { if (locals->PlaneRequiredDISPCLKWithoutODMCombine > MaxMaxDispclkRoundedDown) { - locals->ODMCombineEnablePerState[i][k] = true; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_2to1; mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine; } else if (locals->DSCEnabled[k] && (locals->HActive[k] > DCN21_MAX_DSC_IMAGE_WIDTH)) { - locals->ODMCombineEnablePerState[i][k] = true; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_2to1; mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine; } else if (locals->HActive[k] > DCN21_MAX_420_IMAGE_WIDTH && locals->OutputFormat[k] == dm_420) { - locals->ODMCombineEnablePerState[i][k] = true; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_2to1; mode_lib->vba.PlaneRequiredDISPCLK = mode_lib->vba.PlaneRequiredDISPCLKWithODMCombine; } } @@ -4165,7 +4165,7 @@ void dml21_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l locals->RequiredDISPCLK[i][j] = 0.0; locals->DISPCLK_DPPCLK_Support[i][j] = true; for (k = 0; k <= mode_lib->vba.NumberOfActivePlanes - 1; k++) { - locals->ODMCombineEnablePerState[i][k] = false; + locals->ODMCombineEnablePerState[i][k] = dm_odm_combine_mode_disabled; if (locals->SwathWidthYSingleDPP[k] <= locals->MaximumSwathWidth[k]) { locals->NoOfDPP[i][j][k] = 1; locals->RequiredDPPCLK[i][j][k] = locals->MinDPPCLKUsingSingleDPP[k] @@ -5230,7 +5230,7 @@ void dml21_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l mode_lib->vba.ODMCombineEnabled[k] = locals->ODMCombineEnablePerState[mode_lib->vba.VoltageLevel][k]; } else { - mode_lib->vba.ODMCombineEnabled[k] = false; + mode_lib->vba.ODMCombineEnabled[k] = dm_odm_combine_mode_disabled; } mode_lib->vba.DSCEnabled[k] = locals->RequiresDSC[mode_lib->vba.VoltageLevel][k]; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c index d4c0f9cdac8e..4fa636364793 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c @@ -634,7 +634,7 @@ int dcn30_find_dummy_latency_index_for_fw_based_mclk_switch(struct dc *dc, while (dummy_latency_index < max_latency_table_entries) { context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->dummy_pstate_table[dummy_latency_index].dummy_pstate_latency_us; - dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, false); + dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, false, true); if (context->bw_ctx.dml.soc.allow_dram_self_refresh_or_dram_clock_change_in_vblank == dm_allow_self_refresh_and_mclk_switch) diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c index ec351c8418cb..27f488405335 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c @@ -878,7 +878,9 @@ static bool CalculatePrefetchSchedule( double DSTTotalPixelsAfterScaler; double LineTime; double dst_y_prefetch_equ; +#ifdef __DML_VBA_DEBUG__ double Tsw_oto; +#endif double prefetch_bw_oto; double prefetch_bw_pr; double Tvm_oto; @@ -1060,7 +1062,9 @@ static bool CalculatePrefetchSchedule( min_Lsw = dml_max(1, dml_max(PrefetchSourceLinesY, PrefetchSourceLinesC) / max_vratio_pre); Lsw_oto = dml_ceil(4 * dml_max(prefetch_sw_bytes / prefetch_bw_oto / LineTime, min_Lsw), 1) / 4; +#ifdef __DML_VBA_DEBUG__ Tsw_oto = Lsw_oto * LineTime; +#endif #ifdef __DML_VBA_DEBUG__ diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c index 6a1cf6adea77..acda3e1babd4 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c @@ -149,8 +149,8 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_14_soc = { .num_states = 5, .sr_exit_time_us = 16.5, .sr_enter_plus_exit_time_us = 18.5, - .sr_exit_z8_time_us = 280.0, - .sr_enter_plus_exit_z8_time_us = 350.0, + .sr_exit_z8_time_us = 210.0, + .sr_enter_plus_exit_z8_time_us = 310.0, .writeback_latency_us = 12.0, .dram_channel_width_bytes = 4, .round_trip_ping_latency_dcfclk_cycles = 106, @@ -346,7 +346,8 @@ int dcn314_populate_dml_pipes_from_context_fpu(struct dc *dc, struct dc_state *c context->bw_ctx.dml.ip.det_buffer_size_kbytes = DCN3_14_DEFAULT_DET_SIZE; dc->config.enable_4to1MPC = false; - if (pipe_cnt == 1 && pipe->plane_state && !dc->debug.disable_z9_mpc) { + if (pipe_cnt == 1 && pipe->plane_state + && pipe->plane_state->rotation == ROTATION_ANGLE_0 && !dc->debug.disable_z9_mpc) { if (is_dual_plane(pipe->plane_state->format) && pipe->plane_state->src_rect.width <= 1920 && pipe->plane_state->src_rect.height <= 1080) { dc->config.enable_4to1MPC = true; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c b/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c index cb7c0c878423..c843b394aeb4 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c @@ -900,7 +900,9 @@ static bool CalculatePrefetchSchedule( double DSTTotalPixelsAfterScaler; double LineTime; double dst_y_prefetch_equ; +#ifdef __DML_VBA_DEBUG__ double Tsw_oto; +#endif double prefetch_bw_oto; double prefetch_bw_pr; double Tvm_oto; @@ -1082,7 +1084,9 @@ static bool CalculatePrefetchSchedule( min_Lsw = dml_max(1, dml_max(PrefetchSourceLinesY, PrefetchSourceLinesC) / max_vratio_pre); Lsw_oto = dml_ceil(4 * dml_max(prefetch_sw_bytes / prefetch_bw_oto / LineTime, min_Lsw), 1) / 4; +#ifdef __DML_VBA_DEBUG__ Tsw_oto = Lsw_oto * LineTime; +#endif #ifdef __DML_VBA_DEBUG__ diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_rq_dlg_calc_314.c b/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_rq_dlg_calc_314.c index 61ee9ba063a7..6576b897a512 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_rq_dlg_calc_314.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_rq_dlg_calc_314.c @@ -51,7 +51,7 @@ static bool CalculateBytePerPixelAnd256BBlockSizes( *BytePerPixelDETC = 0; *BytePerPixelY = 4; *BytePerPixelC = 0; - } else if (SourcePixelFormat == dm_444_16 || SourcePixelFormat == dm_444_16) { + } else if (SourcePixelFormat == dm_444_16) { *BytePerPixelDETY = 2; *BytePerPixelDETC = 0; *BytePerPixelY = 2; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c index f94abd124021..e47828e3b6d5 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c @@ -24,13 +24,14 @@ * */ #include "dcn32_fpu.h" -#include "dc_link_dp.h" #include "dcn32/dcn32_resource.h" #include "dcn20/dcn20_resource.h" #include "display_mode_vba_util_32.h" +#include "dml/dcn32/display_mode_vba_32.h" // We need this includes for WATERMARKS_* defines #include "clk_mgr/dcn32/dcn32_smu13_driver_if.h" #include "dcn30/dcn30_resource.h" +#include "link.h" #define DC_LOGGER_INIT(logger) @@ -691,9 +692,11 @@ static bool dcn32_assign_subvp_pipe(struct dc *dc, * to combine this with SubVP can cause issues with the scheduling). * - Not TMZ surface */ - if (pipe->plane_state && !pipe->top_pipe && + if (pipe->plane_state && !pipe->top_pipe && !dcn32_is_center_timing(pipe) && !dcn32_is_psr_capable(pipe) && pipe->stream->mall_stream_config.type == SUBVP_NONE && refresh_rate < 120 && !pipe->plane_state->address.tmz_surface && - vba->ActiveDRAMClockChangeLatencyMarginPerState[vba->VoltageLevel][vba->maxMpcComb][vba->pipe_plane[pipe_idx]] <= 0) { + (vba->ActiveDRAMClockChangeLatencyMarginPerState[vba->VoltageLevel][vba->maxMpcComb][vba->pipe_plane[pipe_idx]] <= 0 || + (vba->ActiveDRAMClockChangeLatencyMarginPerState[vba->VoltageLevel][vba->maxMpcComb][vba->pipe_plane[pipe_idx]] > 0 && + dcn32_allow_subvp_with_active_margin(pipe)))) { while (pipe) { num_pipes++; pipe = pipe->bottom_pipe; @@ -877,6 +880,10 @@ static bool subvp_drr_schedulable(struct dc *dc, struct dc_state *context, struc int16_t stretched_drr_us = 0; int16_t drr_stretched_vblank_us = 0; int16_t max_vblank_mallregion = 0; + const struct dc_config *config = &dc->config; + + if (config->disable_subvp_drr) + return false; // Find SubVP pipe for (i = 0; i < dc->res_pool->pipe_count; i++) { @@ -977,10 +984,12 @@ static bool subvp_vblank_schedulable(struct dc *dc, struct dc_state *context) if (!subvp_pipe && pipe->stream->mall_stream_config.type == SUBVP_MAIN) subvp_pipe = pipe; } - // Use ignore_msa_timing_param flag to identify as DRR - if (found && context->res_ctx.pipe_ctx[vblank_index].stream->ignore_msa_timing_param) { - // SUBVP + DRR case - schedulable = subvp_drr_schedulable(dc, context, &context->res_ctx.pipe_ctx[vblank_index]); + // Use ignore_msa_timing_param and VRR active, or Freesync flag to identify as DRR On + if (found && context->res_ctx.pipe_ctx[vblank_index].stream->ignore_msa_timing_param && + (context->res_ctx.pipe_ctx[vblank_index].stream->allow_freesync || + context->res_ctx.pipe_ctx[vblank_index].stream->vrr_active_variable)) { + // SUBVP + DRR case -- only allowed if run through DRR validation path + schedulable = false; } else if (found) { main_timing = &subvp_pipe->stream->timing; phantom_timing = &subvp_pipe->stream->mall_stream_config.paired_stream->timing; @@ -1084,12 +1093,12 @@ static void dcn32_full_validate_bw_helper(struct dc *dc, { struct vba_vars_st *vba = &context->bw_ctx.dml.vba; unsigned int dc_pipe_idx = 0; + int i = 0; bool found_supported_config = false; struct pipe_ctx *pipe = NULL; uint32_t non_subvp_pipes = 0; bool drr_pipe_found = false; uint32_t drr_pipe_index = 0; - uint32_t i = 0; dc_assert_fp_enabled(); @@ -1169,15 +1178,25 @@ static void dcn32_full_validate_bw_helper(struct dc *dc, pipes[0].clks_cfg.dppclk_mhz = get_dppclk_calculated(&context->bw_ctx.dml, pipes, *pipe_cnt, 0); *vlevel = dml_get_voltage_level(&context->bw_ctx.dml, pipes, *pipe_cnt); + /* Check that vlevel requested supports pstate or not + * if not, select the lowest vlevel that supports it + */ + for (i = *vlevel; i < context->bw_ctx.dml.soc.num_states; i++) { + if (vba->DRAMClockChangeSupport[i][vba->maxMpcComb] != dm_dram_clock_change_unsupported) { + *vlevel = i; + break; + } + } + if (*vlevel < context->bw_ctx.dml.soc.num_states && vba->DRAMClockChangeSupport[*vlevel][vba->maxMpcComb] != dm_dram_clock_change_unsupported && subvp_validate_static_schedulability(dc, context, *vlevel)) { found_supported_config = true; - } else if (*vlevel < context->bw_ctx.dml.soc.num_states && - vba->DRAMClockChangeSupport[*vlevel][vba->maxMpcComb] == dm_dram_clock_change_unsupported) { - /* Case where 1 SubVP is added, and DML reports MCLK unsupported. This handles - * the case for SubVP + DRR, where the DRR display does not support MCLK switch - * at it's native refresh rate / timing. + } else if (*vlevel < context->bw_ctx.dml.soc.num_states) { + /* Case where 1 SubVP is added, and DML reports MCLK unsupported or DRR is allowed. + * This handles the case for SubVP + DRR, where the DRR display does not support MCLK + * switch at it's native refresh rate / timing, or DRR is allowed for the non-subvp + * display. */ for (i = 0; i < dc->res_pool->pipe_count; i++) { pipe = &context->res_ctx.pipe_ctx[i]; @@ -1185,7 +1204,7 @@ static void dcn32_full_validate_bw_helper(struct dc *dc, pipe->stream->mall_stream_config.type == SUBVP_NONE) { non_subvp_pipes++; // Use ignore_msa_timing_param flag to identify as DRR - if (pipe->stream->ignore_msa_timing_param) { + if (pipe->stream->ignore_msa_timing_param && pipe->stream->allow_freesync) { drr_pipe_found = true; drr_pipe_index = i; } @@ -1194,6 +1213,15 @@ static void dcn32_full_validate_bw_helper(struct dc *dc, // If there is only 1 remaining non SubVP pipe that is DRR, check static // schedulability for SubVP + DRR. if (non_subvp_pipes == 1 && drr_pipe_found) { + /* find lowest vlevel that supports the config */ + for (i = *vlevel; i >= 0; i--) { + if (vba->ModeSupport[i][vba->maxMpcComb]) { + *vlevel = i; + } else { + break; + } + } + found_supported_config = subvp_drr_schedulable(dc, context, &context->res_ctx.pipe_ctx[drr_pipe_index]); } @@ -1242,12 +1270,44 @@ static bool is_dtbclk_required(struct dc *dc, struct dc_state *context) for (i = 0; i < dc->res_pool->pipe_count; i++) { if (!context->res_ctx.pipe_ctx[i].stream) continue; - if (is_dp_128b_132b_signal(&context->res_ctx.pipe_ctx[i])) + if (link_is_dp_128b_132b_signal(&context->res_ctx.pipe_ctx[i])) return true; } return false; } +static void dcn20_adjust_freesync_v_startup(const struct dc_crtc_timing *dc_crtc_timing, int *vstartup_start) +{ + struct dc_crtc_timing patched_crtc_timing; + uint32_t asic_blank_end = 0; + uint32_t asic_blank_start = 0; + uint32_t newVstartup = 0; + + patched_crtc_timing = *dc_crtc_timing; + + if (patched_crtc_timing.flags.INTERLACE == 1) { + if (patched_crtc_timing.v_front_porch < 2) + patched_crtc_timing.v_front_porch = 2; + } else { + if (patched_crtc_timing.v_front_porch < 1) + patched_crtc_timing.v_front_porch = 1; + } + + /* blank_start = frame end - front porch */ + asic_blank_start = patched_crtc_timing.v_total - + patched_crtc_timing.v_front_porch; + + /* blank_end = blank_start - active */ + asic_blank_end = asic_blank_start - + patched_crtc_timing.v_border_bottom - + patched_crtc_timing.v_addressable - + patched_crtc_timing.v_border_top; + + newVstartup = asic_blank_end + (patched_crtc_timing.v_total - asic_blank_start); + + *vstartup_start = ((newVstartup > *vstartup_start) ? newVstartup : *vstartup_start); +} + static void dcn32_calculate_dlg_params(struct dc *dc, struct dc_state *context, display_e2e_pipe_params_st *pipes, int pipe_cnt, int vlevel) @@ -1270,7 +1330,6 @@ static void dcn32_calculate_dlg_params(struct dc *dc, struct dc_state *context, context->bw_ctx.bw.dcn.clk.p_state_change_support = context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][context->bw_ctx.dml.vba.maxMpcComb] != dm_dram_clock_change_unsupported; - context->bw_ctx.bw.dcn.clk.num_ways = dcn32_helper_calculate_num_ways_for_subvp(dc, context); context->bw_ctx.bw.dcn.clk.dppclk_khz = 0; context->bw_ctx.bw.dcn.clk.dtbclk_en = is_dtbclk_required(dc, context); @@ -1294,6 +1353,10 @@ static void dcn32_calculate_dlg_params(struct dc *dc, struct dc_state *context, unbounded_req_enabled = false; } + context->bw_ctx.bw.dcn.mall_ss_size_bytes = 0; + context->bw_ctx.bw.dcn.mall_ss_psr_active_size_bytes = 0; + context->bw_ctx.bw.dcn.mall_subvp_size_bytes = 0; + for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) { if (!context->res_ctx.pipe_ctx[i].stream) continue; @@ -1325,6 +1388,34 @@ static void dcn32_calculate_dlg_params(struct dc *dc, struct dc_state *context, else context->res_ctx.pipe_ctx[i].plane_res.bw.dppclk_khz = 0; context->res_ctx.pipe_ctx[i].pipe_dlg_param = pipes[pipe_idx].pipe.dest; + + context->res_ctx.pipe_ctx[i].surface_size_in_mall_bytes = get_surface_size_in_mall(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx); + + /* MALL Allocation Sizes */ + /* count from active, top pipes per plane only */ + if (context->res_ctx.pipe_ctx[i].stream && context->res_ctx.pipe_ctx[i].plane_state && + (context->res_ctx.pipe_ctx[i].top_pipe == NULL || + context->res_ctx.pipe_ctx[i].plane_state != context->res_ctx.pipe_ctx[i].top_pipe->plane_state) && + context->res_ctx.pipe_ctx[i].prev_odm_pipe == NULL) { + /* SS: all active surfaces stored in MALL */ + if (context->res_ctx.pipe_ctx[i].stream->mall_stream_config.type != SUBVP_PHANTOM) { + context->bw_ctx.bw.dcn.mall_ss_size_bytes += context->res_ctx.pipe_ctx[i].surface_size_in_mall_bytes; + + if (context->res_ctx.pipe_ctx[i].stream->link->psr_settings.psr_version == DC_PSR_VERSION_UNSUPPORTED) { + /* SS PSR On: all active surfaces part of streams not supporting PSR stored in MALL */ + context->bw_ctx.bw.dcn.mall_ss_psr_active_size_bytes += context->res_ctx.pipe_ctx[i].surface_size_in_mall_bytes; + } + } else { + /* SUBVP: phantom surfaces only stored in MALL */ + context->bw_ctx.bw.dcn.mall_subvp_size_bytes += context->res_ctx.pipe_ctx[i].surface_size_in_mall_bytes; + } + } + + if (context->res_ctx.pipe_ctx[i].stream->adaptive_sync_infopacket.valid) + dcn20_adjust_freesync_v_startup( + &context->res_ctx.pipe_ctx[i].stream->timing, + &context->res_ctx.pipe_ctx[i].pipe_dlg_param.vstartup_start); + pipe_idx++; } /* If DCN isn't making memory requests we can allow pstate change and lower clocks */ @@ -1345,6 +1436,8 @@ static void dcn32_calculate_dlg_params(struct dc *dc, struct dc_state *context, context->bw_ctx.bw.dcn.clk.max_supported_dispclk_khz = context->bw_ctx.dml.soc.clock_limits[vlevel].dispclk_mhz * 1000; + context->bw_ctx.bw.dcn.clk.num_ways = dcn32_helper_calculate_num_ways_for_subvp(dc, context); + context->bw_ctx.bw.dcn.compbuf_size_kb = context->bw_ctx.dml.ip.config_return_buffer_size_in_kbytes; for (i = 0; i < dc->res_pool->pipe_count; i++) { @@ -1530,6 +1623,7 @@ bool dcn32_internal_validate_bw(struct dc *dc, } dml_log_pipe_params(&context->bw_ctx.dml, pipes, pipe_cnt); + context->bw_ctx.dml.soc.max_vratio_pre = dcn32_determine_max_vratio_prefetch(dc, context); if (!fast_validate) dcn32_full_validate_bw_helper(dc, context, pipes, &vlevel, split, merge, &pipe_cnt); @@ -1549,16 +1643,12 @@ bool dcn32_internal_validate_bw(struct dc *dc, * to support with Prefetch mode 1 (dm_prefetch_support_fclk_and_stutter == 2) */ context->bw_ctx.dml.soc.allow_for_pstate_or_stutter_in_vblank_final = - dm_prefetch_support_fclk_and_stutter; + dm_prefetch_support_none; + context->bw_ctx.dml.validate_max_state = fast_validate; vlevel = dml_get_voltage_level(&context->bw_ctx.dml, pipes, pipe_cnt); - /* Last attempt with Prefetch mode 2 (dm_prefetch_support_stutter == 3) */ - if (vlevel == context->bw_ctx.dml.soc.num_states) { - context->bw_ctx.dml.soc.allow_for_pstate_or_stutter_in_vblank_final = - dm_prefetch_support_stutter; - vlevel = dml_get_voltage_level(&context->bw_ctx.dml, pipes, pipe_cnt); - } + context->bw_ctx.dml.validate_max_state = false; if (vlevel < context->bw_ctx.dml.soc.num_states) { memset(split, 0, sizeof(split)); @@ -1645,6 +1735,7 @@ bool dcn32_internal_validate_bw(struct dc *dc, dcn20_release_dsc(&context->res_ctx, dc->res_pool, &pipe->stream_res.dsc); memset(&pipe->plane_res, 0, sizeof(pipe->plane_res)); memset(&pipe->stream_res, 0, sizeof(pipe->stream_res)); + memset(&pipe->link_res, 0, sizeof(pipe->link_res)); repopulate_pipes = true; } else if (pipe->top_pipe && pipe->top_pipe->plane_state == pipe->plane_state) { struct pipe_ctx *top_pipe = pipe->top_pipe; @@ -1660,6 +1751,7 @@ bool dcn32_internal_validate_bw(struct dc *dc, pipe->stream = NULL; memset(&pipe->plane_res, 0, sizeof(pipe->plane_res)); memset(&pipe->stream_res, 0, sizeof(pipe->stream_res)); + memset(&pipe->link_res, 0, sizeof(pipe->link_res)); repopulate_pipes = true; } else ASSERT(0); /* Should never try to merge master pipe */ @@ -1834,7 +1926,7 @@ void dcn32_calculate_wm_and_dlg_fpu(struct dc *dc, struct dc_state *context, bool subvp_in_use = dcn32_subvp_in_use(dc, context); unsigned int min_dram_speed_mts_margin; bool need_fclk_lat_as_dummy = false; - bool is_subvp_p_drr = true; + bool is_subvp_p_drr = false; dc_assert_fp_enabled(); @@ -1842,7 +1934,8 @@ void dcn32_calculate_wm_and_dlg_fpu(struct dc *dc, struct dc_state *context, if (subvp_in_use) { /* Override DRAMClockChangeSupport for SubVP + DRR case where the DRR cannot switch without stretching it's VBLANK */ if (!pstate_en) { - context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][context->bw_ctx.dml.vba.maxMpcComb] = dm_dram_clock_change_vblank_w_mall_sub_vp; + context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][maxMpcComb] = dm_dram_clock_change_vblank_w_mall_sub_vp; + context->bw_ctx.dml.soc.allow_for_pstate_or_stutter_in_vblank_final = dm_prefetch_support_fclk_and_stutter; pstate_en = true; is_subvp_p_drr = true; } @@ -1860,8 +1953,9 @@ void dcn32_calculate_wm_and_dlg_fpu(struct dc *dc, struct dc_state *context, context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.pstate_latency_us; dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, false); + maxMpcComb = context->bw_ctx.dml.vba.maxMpcComb; if (is_subvp_p_drr) { - context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][context->bw_ctx.dml.vba.maxMpcComb] = dm_dram_clock_change_vblank_w_mall_sub_vp; + context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][maxMpcComb] = dm_dram_clock_change_vblank_w_mall_sub_vp; } } @@ -2038,6 +2132,10 @@ void dcn32_calculate_wm_and_dlg_fpu(struct dc *dc, struct dc_state *context, */ context->bw_ctx.bw.dcn.watermarks.a = context->bw_ctx.bw.dcn.watermarks.c; context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns = 0; + /* Calculate FCLK p-state change watermark based on FCLK pstate change latency in case + * UCLK p-state is not supported, to avoid underflow in case FCLK pstate is supported + */ + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.fclk_pstate_change_ns = get_fclk_watermark(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; } else { /* Set A: * All clocks min. @@ -2443,8 +2541,11 @@ void dcn32_update_bw_bounding_box_fpu(struct dc *dc, struct clk_bw_params *bw_pa } /* Override from VBIOS for num_chan */ - if (dc->ctx->dc_bios->vram_info.num_chans) + if (dc->ctx->dc_bios->vram_info.num_chans) { dcn3_2_soc.num_chans = dc->ctx->dc_bios->vram_info.num_chans; + dcn3_2_soc.mall_allocated_for_dcn_mbytes = (double)(dcn32_calc_num_avail_chans_for_mall(dc, + dc->ctx->dc_bios->vram_info.num_chans) * dc->caps.mall_size_per_mem_channel); + } if (dc->ctx->dc_bios->vram_info.dram_channel_width_bytes) dcn3_2_soc.dram_channel_width_bytes = dc->ctx->dc_bios->vram_info.dram_channel_width_bytes; @@ -2622,3 +2723,60 @@ void dcn32_zero_pipe_dcc_fraction(display_e2e_pipe_params_st *pipes, pipes[pipe_cnt].pipe.src.dcc_fraction_of_zs_req_luma = 0; pipes[pipe_cnt].pipe.src.dcc_fraction_of_zs_req_chroma = 0; } + +bool dcn32_allow_subvp_with_active_margin(struct pipe_ctx *pipe) +{ + bool allow = false; + uint32_t refresh_rate = 0; + + /* Allow subvp on displays that have active margin for 2560x1440@60hz displays + * only for now. There must be no scaling as well. + * + * For now we only enable on 2560x1440@60hz displays to enable 4K60 + 1440p60 configs + * for p-state switching. + */ + if (pipe->stream && pipe->plane_state) { + refresh_rate = (pipe->stream->timing.pix_clk_100hz * 100 + + pipe->stream->timing.v_total * pipe->stream->timing.h_total - 1) + / (double)(pipe->stream->timing.v_total * pipe->stream->timing.h_total); + if (pipe->stream->timing.v_addressable == 1440 && + pipe->stream->timing.h_addressable == 2560 && + refresh_rate >= 55 && refresh_rate <= 65 && + pipe->plane_state->src_rect.height == 1440 && + pipe->plane_state->src_rect.width == 2560 && + pipe->plane_state->dst_rect.height == 1440 && + pipe->plane_state->dst_rect.width == 2560) + allow = true; + } + return allow; +} + +/** + * ******************************************************************************************* + * dcn32_determine_max_vratio_prefetch: Determine max Vratio for prefetch by driver policy + * + * @param [in]: dc: Current DC state + * @param [in]: context: New DC state to be programmed + * + * @return: Max vratio for prefetch + * + * ******************************************************************************************* + */ +double dcn32_determine_max_vratio_prefetch(struct dc *dc, struct dc_state *context) +{ + double max_vratio_pre = __DML_MAX_BW_RATIO_PRE__; // Default value is 4 + int i; + + /* For single display MPO configs, allow the max vratio to be 8 + * if any plane is YUV420 format + */ + if (context->stream_count == 1 && context->stream_status[0].plane_count > 1) { + for (i = 0; i < context->stream_status[0].plane_count; i++) { + if (context->stream_status[0].plane_states[i]->format == SURFACE_PIXEL_FORMAT_VIDEO_420_YCbCr || + context->stream_status[0].plane_states[i]->format == SURFACE_PIXEL_FORMAT_VIDEO_420_YCrCb) { + max_vratio_pre = __DML_MAX_VRATIO_PRE__; + } + } + } + return max_vratio_pre; +} diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.c index 4b8f5fa0f0ad..3b2a014ccf8f 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.c @@ -387,6 +387,7 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman mode_lib->vba.NumberOfActiveSurfaces, mode_lib->vba.MALLAllocatedForDCNFinal, mode_lib->vba.UseMALLForStaticScreen, + mode_lib->vba.UsesMALLForPStateChange, mode_lib->vba.DCCEnable, mode_lib->vba.ViewportStationary, mode_lib->vba.ViewportXStartY, @@ -411,6 +412,8 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman v->BlockWidthC, v->BlockHeightY, v->BlockHeightC, + mode_lib->vba.DCCMetaPitchY, + mode_lib->vba.DCCMetaPitchC, /* Output */ v->SurfaceSizeInMALL, @@ -893,8 +896,8 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman if (v->DestinationLinesForPrefetch[k] < 2) DestinationLineTimesForPrefetchLessThan2 = true; - if (v->VRatioPrefetchY[k] > __DML_MAX_VRATIO_PRE__ - || v->VRatioPrefetchC[k] > __DML_MAX_VRATIO_PRE__) + if (v->VRatioPrefetchY[k] > v->MaxVRatioPre + || v->VRatioPrefetchC[k] > v->MaxVRatioPre) VRatioPrefetchMoreThanMax = true; //bool DestinationLinesToRequestVMInVBlankEqualOrMoreThan32 = false; @@ -939,6 +942,9 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman v->UrgBurstFactorLumaPre, v->UrgBurstFactorChromaPre, v->UrgBurstFactorCursorPre, + v->PrefetchBandwidth, + v->VRatio, + v->MaxVRatioPre, /* output */ &MaxTotalRDBandwidth, @@ -969,6 +975,9 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman v->dummy_vars.DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation.dummy_unit_vector, v->dummy_vars.DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation.dummy_unit_vector, v->dummy_vars.DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation.dummy_unit_vector, + v->PrefetchBandwidth, + v->VRatio, + v->MaxVRatioPre, /* output */ &v->dummy_vars.DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerformanceCalculation.dummy_single[0], @@ -1636,9 +1645,14 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman static void mode_support_configuration(struct vba_vars_st *v, struct display_mode_lib *mode_lib) { - int i, j; + int i, j, start_state; + + if (mode_lib->validate_max_state) + start_state = v->soc.num_states - 1; + else + start_state = 0; - for (i = v->soc.num_states - 1; i >= 0; i--) { + for (i = v->soc.num_states - 1; i >= start_state; i--) { for (j = 0; j < 2; j++) { if (mode_lib->vba.ScaleRatioAndTapsSupport == true && mode_lib->vba.SourceFormatPixelAndScanSupport == true @@ -1707,7 +1721,7 @@ static void mode_support_configuration(struct vba_vars_st *v, void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_lib) { struct vba_vars_st *v = &mode_lib->vba; - int i, j; + int i, j, start_state; unsigned int k, m; unsigned int MaximumMPCCombine; unsigned int NumberOfNonCombinedSurfaceOfMaximumBandwidth; @@ -1720,6 +1734,10 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l #endif /*MODE SUPPORT, VOLTAGE STATE AND SOC CONFIGURATION*/ + if (mode_lib->validate_max_state) + start_state = v->soc.num_states - 1; + else + start_state = 0; /*Scale Ratio, taps Support Check*/ @@ -2009,7 +2027,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l mode_lib->vba.MPCCombineMethodIncompatible = v->dummy_vars.dml32_ModeSupportAndSystemConfigurationFull.MPCCombineMethodAsNeededForPStateChangeAndVoltage && v->dummy_vars.dml32_ModeSupportAndSystemConfigurationFull.MPCCombineMethodAsPossible; - for (i = 0; i < v->soc.num_states; i++) { + for (i = start_state; i < v->soc.num_states; i++) { for (j = 0; j < 2; j++) { mode_lib->vba.TotalNumberOfActiveDPP[i][j] = 0; mode_lib->vba.TotalAvailablePipesSupport[i][j] = true; @@ -2286,7 +2304,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l } } - for (i = 0; i < v->soc.num_states; ++i) { + for (i = start_state; i < v->soc.num_states; ++i) { mode_lib->vba.ExceededMultistreamSlots[i] = false; for (k = 0; k < mode_lib->vba.NumberOfActiveSurfaces; ++k) { if (mode_lib->vba.OutputMultistreamEn[k] == true && mode_lib->vba.OutputMultistreamId[k] == k) { @@ -2335,8 +2353,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l if (mode_lib->vba.DSCEnable[k] && mode_lib->vba.ForcedOutputLinkBPP[k] != 0) mode_lib->vba.DSCOnlyIfNecessaryWithBPP = true; - if ((mode_lib->vba.DSCEnable[k] || mode_lib->vba.DSCEnable[k]) - && mode_lib->vba.OutputFormat[k] == dm_n422 + if (mode_lib->vba.DSCEnable[k] && mode_lib->vba.OutputFormat[k] == dm_n422 && !mode_lib->vba.DSC422NativeSupport) mode_lib->vba.DSC422NativeNotSupported = true; @@ -2386,7 +2403,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l } } - for (i = 0; i < v->soc.num_states; ++i) { + for (i = start_state; i < v->soc.num_states; ++i) { mode_lib->vba.DTBCLKRequiredMoreThanSupported[i] = false; for (k = 0; k < mode_lib->vba.NumberOfActiveSurfaces; ++k) { if (mode_lib->vba.BlendingAndTiming[k] == k @@ -2403,7 +2420,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l } } - for (i = 0; i < v->soc.num_states; ++i) { + for (i = start_state; i < v->soc.num_states; ++i) { mode_lib->vba.ODMCombine2To1SupportCheckOK[i] = true; mode_lib->vba.ODMCombine4To1SupportCheckOK[i] = true; for (k = 0; k < mode_lib->vba.NumberOfActiveSurfaces; ++k) { @@ -2421,7 +2438,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l } } - for (i = 0; i < v->soc.num_states; i++) { + for (i = start_state; i < v->soc.num_states; i++) { mode_lib->vba.DSCCLKRequiredMoreThanSupported[i] = false; for (k = 0; k <= mode_lib->vba.NumberOfActiveSurfaces - 1; k++) { if (mode_lib->vba.BlendingAndTiming[k] == k) { @@ -2458,7 +2475,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l /* Check DSC Unit and Slices Support */ v->dummy_vars.dml32_ModeSupportAndSystemConfigurationFull.TotalDSCUnitsRequired = 0; - for (i = 0; i < v->soc.num_states; ++i) { + for (i = start_state; i < v->soc.num_states; ++i) { mode_lib->vba.NotEnoughDSCUnits[i] = false; mode_lib->vba.NotEnoughDSCSlices[i] = false; v->dummy_vars.dml32_ModeSupportAndSystemConfigurationFull.TotalDSCUnitsRequired = 0; @@ -2493,7 +2510,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l } /*DSC Delay per state*/ - for (i = 0; i < v->soc.num_states; ++i) { + for (i = start_state; i < v->soc.num_states; ++i) { for (k = 0; k < mode_lib->vba.NumberOfActiveSurfaces; ++k) { mode_lib->vba.DSCDelayPerState[i][k] = dml32_DSCDelayRequirement( mode_lib->vba.RequiresDSC[i][k], mode_lib->vba.ODMCombineEnablePerState[i][k], @@ -2520,7 +2537,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l //Calculate Swath, DET Configuration, DCFCLKDeepSleep // - for (i = 0; i < (int) v->soc.num_states; ++i) { + for (i = start_state; i < (int) v->soc.num_states; ++i) { for (j = 0; j <= 1; ++j) { for (k = 0; k < mode_lib->vba.NumberOfActiveSurfaces; ++k) { mode_lib->vba.RequiredDPPCLKThisState[k] = mode_lib->vba.RequiredDPPCLK[i][j][k]; @@ -2626,6 +2643,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l mode_lib->vba.NumberOfActiveSurfaces, mode_lib->vba.MALLAllocatedForDCNFinal, mode_lib->vba.UseMALLForStaticScreen, + mode_lib->vba.UsesMALLForPStateChange, mode_lib->vba.DCCEnable, mode_lib->vba.ViewportStationary, mode_lib->vba.ViewportXStartY, @@ -2650,12 +2668,14 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l mode_lib->vba.MacroTileWidthC, mode_lib->vba.MacroTileHeightY, mode_lib->vba.MacroTileHeightC, + mode_lib->vba.DCCMetaPitchY, + mode_lib->vba.DCCMetaPitchC, /* Output */ mode_lib->vba.SurfaceSizeInMALL, &mode_lib->vba.ExceededMALLSize); - for (i = 0; i < v->soc.num_states; i++) { + for (i = start_state; i < v->soc.num_states; i++) { for (j = 0; j < 2; j++) { for (k = 0; k <= mode_lib->vba.NumberOfActiveSurfaces - 1; k++) { mode_lib->vba.swath_width_luma_ub_this_state[k] = @@ -2882,7 +2902,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l } //Calculate Return BW - for (i = 0; i < (int) v->soc.num_states; ++i) { + for (i = start_state; i < (int) v->soc.num_states; ++i) { for (j = 0; j <= 1; ++j) { for (k = 0; k <= mode_lib->vba.NumberOfActiveSurfaces - 1; k++) { if (mode_lib->vba.BlendingAndTiming[k] == k) { @@ -2961,7 +2981,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l &mode_lib->vba.MinPrefetchMode, &mode_lib->vba.MaxPrefetchMode); - for (i = 0; i < (int) v->soc.num_states; ++i) { + for (i = start_state; i < (int) v->soc.num_states; ++i) { for (j = 0; j <= 1; ++j) mode_lib->vba.DCFCLKState[i][j] = mode_lib->vba.DCFCLKPerState[i]; } @@ -3083,7 +3103,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l mode_lib->vba.DCFCLKState); } // UseMinimumRequiredDCFCLK == true - for (i = 0; i < (int) v->soc.num_states; ++i) { + for (i = start_state; i < (int) v->soc.num_states; ++i) { for (j = 0; j <= 1; ++j) { mode_lib->vba.ReturnBWPerState[i][j] = dml32_get_return_bw_mbps(&mode_lib->vba.soc, i, mode_lib->vba.HostVMEnable, mode_lib->vba.DCFCLKState[i][j], @@ -3092,7 +3112,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l } //Re-ordering Buffer Support Check - for (i = 0; i < (int) v->soc.num_states; ++i) { + for (i = start_state; i < (int) v->soc.num_states; ++i) { for (j = 0; j <= 1; ++j) { if ((mode_lib->vba.ROBBufferSizeInKByte - mode_lib->vba.PixelChunkSizeInKByte) * 1024 / mode_lib->vba.ReturnBWPerState[i][j] @@ -3114,7 +3134,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l + mode_lib->vba.ReadBandwidthChroma[k]; } - for (i = 0; i < (int) v->soc.num_states; ++i) { + for (i = start_state; i < (int) v->soc.num_states; ++i) { for (j = 0; j <= 1; ++j) { mode_lib->vba.MaxTotalVerticalActiveAvailableBandwidth[i][j] = dml_min3(mode_lib->vba.ReturnBusWidth * mode_lib->vba.DCFCLKState[i][j] @@ -3138,7 +3158,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l /* Prefetch Check */ - for (i = 0; i < (int) v->soc.num_states; ++i) { + for (i = start_state; i < (int) v->soc.num_states; ++i) { for (j = 0; j <= 1; ++j) { mode_lib->vba.TimeCalc = 24 / mode_lib->vba.ProjectedDCFCLKDeepSleep[i][j]; @@ -3358,6 +3378,9 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l mode_lib->vba.UrgentBurstFactorLumaPre, mode_lib->vba.UrgentBurstFactorChromaPre, mode_lib->vba.UrgentBurstFactorCursorPre, + v->PrefetchBW, + v->VRatio, + v->MaxVRatioPre, /* output */ &v->dummy_vars.dml32_ModeSupportAndSystemConfigurationFull.dummy_single[0], // Single *PrefetchBandwidth @@ -3382,8 +3405,8 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l mode_lib->vba.VRatioInPrefetchSupported[i][j] = true; for (k = 0; k <= mode_lib->vba.NumberOfActiveSurfaces - 1; k++) { - if (mode_lib->vba.VRatioPreY[i][j][k] > __DML_MAX_VRATIO_PRE__ - || mode_lib->vba.VRatioPreC[i][j][k] > __DML_MAX_VRATIO_PRE__ + if (mode_lib->vba.VRatioPreY[i][j][k] > mode_lib->vba.MaxVRatioPre + || mode_lib->vba.VRatioPreC[i][j][k] > mode_lib->vba.MaxVRatioPre || mode_lib->vba.NoTimeForPrefetch[i][j][k] == true) { mode_lib->vba.VRatioInPrefetchSupported[i][j] = false; } @@ -3639,7 +3662,6 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l if (mode_lib->vba.SourcePixelFormat[k] != dm_444_64 && mode_lib->vba.SourcePixelFormat[k] != dm_444_32 && mode_lib->vba.SourcePixelFormat[k] != dm_444_16 - && mode_lib->vba.SourcePixelFormat[k] != dm_444_16 && mode_lib->vba.SourcePixelFormat[k] != dm_444_8 && mode_lib->vba.SourcePixelFormat[k] != dm_rgbe) { if (mode_lib->vba.ViewportWidthChroma[k] > mode_lib->vba.SurfaceWidthC[k] @@ -3656,7 +3678,7 @@ void dml32_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l MaximumMPCCombine = 0; - for (i = v->soc.num_states; i >= 0; i--) { + for (i = v->soc.num_states; i >= start_state; i--) { if (i == v->soc.num_states || mode_lib->vba.ModeSupport[i][0] == true || mode_lib->vba.ModeSupport[i][1] == true) { mode_lib->vba.VoltageLevel = i; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.h b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.h index c8b28c83ddf4..500b3dd6052d 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_32.h @@ -44,7 +44,8 @@ #define __DML_MIN_DCFCLK_FACTOR__ 1.15 // Prefetch schedule max vratio -#define __DML_MAX_VRATIO_PRE__ 4.0 +#define __DML_MAX_VRATIO_PRE__ 7.9 +#define __DML_MAX_BW_RATIO_PRE__ 4.0 #define __DML_VBA_MAX_DST_Y_PRE__ 63.75 diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c index b53feeaf5cf1..d1000aa4c481 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c @@ -1772,6 +1772,7 @@ void dml32_CalculateSurfaceSizeInMall( unsigned int NumberOfActiveSurfaces, unsigned int MALLAllocatedForDCN, enum dm_use_mall_for_static_screen_mode UseMALLForStaticScreen[], + enum dm_use_mall_for_pstate_change_mode UsesMALLForPStateChange[], bool DCCEnable[], bool ViewportStationary[], unsigned int ViewportXStartY[], @@ -1796,13 +1797,17 @@ void dml32_CalculateSurfaceSizeInMall( unsigned int ReadBlockWidthC[], unsigned int ReadBlockHeightY[], unsigned int ReadBlockHeightC[], + unsigned int DCCMetaPitchY[], + unsigned int DCCMetaPitchC[], /* Output */ unsigned int SurfaceSizeInMALL[], bool *ExceededMALLSize) { - unsigned int TotalSurfaceSizeInMALL = 0; unsigned int k; + unsigned int TotalSurfaceSizeInMALLForSS = 0; + unsigned int TotalSurfaceSizeInMALLForSubVP = 0; + unsigned int MALLAllocatedForDCNInBytes = MALLAllocatedForDCN * 1024 * 1024; for (k = 0; k < NumberOfActiveSurfaces; ++k) { if (ViewportStationary[k]) { @@ -1828,18 +1833,18 @@ void dml32_CalculateSurfaceSizeInMall( } if (DCCEnable[k] == true) { SurfaceSizeInMALL[k] = SurfaceSizeInMALL[k] + - dml_min(dml_ceil(SurfaceWidthY[k], 8 * Read256BytesBlockWidthY[k]), + (dml_min(dml_ceil(DCCMetaPitchY[k], 8 * Read256BytesBlockWidthY[k]), dml_floor(ViewportXStartY[k] + ViewportWidthY[k] + 8 * Read256BytesBlockWidthY[k] - 1, 8 * Read256BytesBlockWidthY[k]) - dml_floor(ViewportXStartY[k], 8 * Read256BytesBlockWidthY[k])) * dml_min(dml_ceil(SurfaceHeightY[k], 8 * Read256BytesBlockHeightY[k]), dml_floor(ViewportYStartY[k] + ViewportHeightY[k] + 8 * Read256BytesBlockHeightY[k] - 1, 8 * - Read256BytesBlockHeightY[k]) - dml_floor(ViewportYStartY[k], 8 - * Read256BytesBlockHeightY[k])) * BytesPerPixelY[k] / 256; + Read256BytesBlockHeightY[k]) - dml_floor(ViewportYStartY[k], 8 * + Read256BytesBlockHeightY[k])) * BytesPerPixelY[k] / 256) + (64 * 1024); if (Read256BytesBlockWidthC[k] > 0) { SurfaceSizeInMALL[k] = SurfaceSizeInMALL[k] + - dml_min(dml_ceil(SurfaceWidthC[k], 8 * + dml_min(dml_ceil(DCCMetaPitchC[k], 8 * Read256BytesBlockWidthC[k]), dml_floor(ViewportXStartC[k] + ViewportWidthC[k] + 8 * Read256BytesBlockWidthC[k] - 1, 8 * @@ -1872,16 +1877,16 @@ void dml32_CalculateSurfaceSizeInMall( } if (DCCEnable[k] == true) { SurfaceSizeInMALL[k] = SurfaceSizeInMALL[k] + - dml_ceil(dml_min(SurfaceWidthY[k], ViewportWidthY[k] + 8 * + (dml_ceil(dml_min(DCCMetaPitchY[k], ViewportWidthY[k] + 8 * Read256BytesBlockWidthY[k] - 1), 8 * Read256BytesBlockWidthY[k]) * dml_ceil(dml_min(SurfaceHeightY[k], ViewportHeightY[k] + 8 * Read256BytesBlockHeightY[k] - 1), 8 * - Read256BytesBlockHeightY[k]) * BytesPerPixelY[k] / 256; + Read256BytesBlockHeightY[k]) * BytesPerPixelY[k] / 256) + (64 * 1024); if (Read256BytesBlockWidthC[k] > 0) { SurfaceSizeInMALL[k] = SurfaceSizeInMALL[k] + - dml_ceil(dml_min(SurfaceWidthC[k], ViewportWidthC[k] + 8 * + dml_ceil(dml_min(DCCMetaPitchC[k], ViewportWidthC[k] + 8 * Read256BytesBlockWidthC[k] - 1), 8 * Read256BytesBlockWidthC[k]) * dml_ceil(dml_min(SurfaceHeightC[k], ViewportHeightC[k] + 8 * @@ -1894,10 +1899,14 @@ void dml32_CalculateSurfaceSizeInMall( } for (k = 0; k < NumberOfActiveSurfaces; ++k) { - if (UseMALLForStaticScreen[k] == dm_use_mall_static_screen_enable) - TotalSurfaceSizeInMALL = TotalSurfaceSizeInMALL + SurfaceSizeInMALL[k]; + /* SS and Subvp counted separate as they are never used at the same time */ + if (UsesMALLForPStateChange[k] == dm_use_mall_pstate_change_phantom_pipe) + TotalSurfaceSizeInMALLForSubVP = TotalSurfaceSizeInMALLForSubVP + SurfaceSizeInMALL[k]; + else if (UseMALLForStaticScreen[k] == dm_use_mall_static_screen_enable) + TotalSurfaceSizeInMALLForSS = TotalSurfaceSizeInMALLForSS + SurfaceSizeInMALL[k]; } - *ExceededMALLSize = (TotalSurfaceSizeInMALL > MALLAllocatedForDCN * 1024 * 1024); + *ExceededMALLSize = (TotalSurfaceSizeInMALLForSS > MALLAllocatedForDCNInBytes) || + (TotalSurfaceSizeInMALLForSubVP > MALLAllocatedForDCNInBytes); } // CalculateSurfaceSizeInMall void dml32_CalculateVMRowAndSwath( @@ -3471,7 +3480,7 @@ bool dml32_CalculatePrefetchSchedule( double prefetch_sw_bytes; double bytes_pp; double dep_bytes; - unsigned int max_vratio_pre = __DML_MAX_VRATIO_PRE__; + unsigned int max_vratio_pre = v->MaxVRatioPre; double min_Lsw; double Tsw_est1 = 0; double Tsw_est3 = 0; @@ -6134,29 +6143,46 @@ void dml32_CalculatePrefetchBandwithSupport(unsigned int NumberOfActiveSurfaces, double UrgentBurstFactorLumaPre[], double UrgentBurstFactorChromaPre[], double UrgentBurstFactorCursorPre[], + double PrefetchBW[], + double VRatio[], + double MaxVRatioPre, /* output */ - double *PrefetchBandwidth, + double *MaxPrefetchBandwidth, double *FractionOfUrgentBandwidth, bool *PrefetchBandwidthSupport) { unsigned int k; + double ActiveBandwidthPerSurface; bool NotEnoughUrgentLatencyHiding = false; + double TotalActiveBandwidth = 0; + double TotalPrefetchBandwidth = 0; + for (k = 0; k < NumberOfActiveSurfaces; ++k) { if (NotUrgentLatencyHiding[k]) { NotEnoughUrgentLatencyHiding = true; } } - *PrefetchBandwidth = 0; + *MaxPrefetchBandwidth = 0; for (k = 0; k < NumberOfActiveSurfaces; ++k) { - *PrefetchBandwidth = *PrefetchBandwidth + dml_max3(NumberOfDPP[k] * prefetch_vmrow_bw[k], - ReadBandwidthLuma[k] * UrgentBurstFactorLuma[k] + ReadBandwidthChroma[k] * UrgentBurstFactorChroma[k] + cursor_bw[k] * UrgentBurstFactorCursor[k] + NumberOfDPP[k] * (meta_row_bandwidth[k] + dpte_row_bandwidth[k]), + ActiveBandwidthPerSurface = ReadBandwidthLuma[k] * UrgentBurstFactorLuma[k] + ReadBandwidthChroma[k] * UrgentBurstFactorChroma[k] + cursor_bw[k] * UrgentBurstFactorCursor[k] + NumberOfDPP[k] * (meta_row_bandwidth[k] + dpte_row_bandwidth[k]); + + TotalActiveBandwidth += ActiveBandwidthPerSurface; + + TotalPrefetchBandwidth = TotalPrefetchBandwidth + PrefetchBW[k] * VRatio[k]; + + *MaxPrefetchBandwidth = *MaxPrefetchBandwidth + dml_max3(NumberOfDPP[k] * prefetch_vmrow_bw[k], + ActiveBandwidthPerSurface, NumberOfDPP[k] * (PrefetchBandwidthLuma[k] * UrgentBurstFactorLumaPre[k] + PrefetchBandwidthChroma[k] * UrgentBurstFactorChromaPre[k]) + cursor_bw_pre[k] * UrgentBurstFactorCursorPre[k]); } - *PrefetchBandwidthSupport = (*PrefetchBandwidth <= ReturnBW) && !NotEnoughUrgentLatencyHiding; - *FractionOfUrgentBandwidth = *PrefetchBandwidth / ReturnBW; + if (MaxVRatioPre == __DML_MAX_VRATIO_PRE__) + *PrefetchBandwidthSupport = (*MaxPrefetchBandwidth <= ReturnBW) && (TotalPrefetchBandwidth <= TotalActiveBandwidth * __DML_MAX_BW_RATIO_PRE__) && !NotEnoughUrgentLatencyHiding; + else + *PrefetchBandwidthSupport = (*MaxPrefetchBandwidth <= ReturnBW) && !NotEnoughUrgentLatencyHiding; + + *FractionOfUrgentBandwidth = *MaxPrefetchBandwidth / ReturnBW; } double dml32_CalculateBandwidthAvailableForImmediateFlip(unsigned int NumberOfActiveSurfaces, @@ -6245,7 +6271,7 @@ bool dml32_CalculateDETSwathFillLatencyHiding(unsigned int NumberOfActiveSurface double PixelClock[], double VRatioY[], double VRatioC[], - enum dm_use_mall_for_pstate_change_mode UsesMALLForPStateChange[DC__NUM_DPP__MAX]) + enum dm_use_mall_for_pstate_change_mode UsesMALLForPStateChange[]) { int k; double SwathSizeAllSurfaces = 0; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.h b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.h index 779c6805f599..9ba792c633a5 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.h @@ -334,6 +334,7 @@ void dml32_CalculateSurfaceSizeInMall( unsigned int NumberOfActiveSurfaces, unsigned int MALLAllocatedForDCN, enum dm_use_mall_for_static_screen_mode UseMALLForStaticScreen[], + enum dm_use_mall_for_pstate_change_mode UsesMALLForPStateChange[], bool DCCEnable[], bool ViewportStationary[], unsigned int ViewportXStartY[], @@ -358,6 +359,8 @@ void dml32_CalculateSurfaceSizeInMall( unsigned int ReadBlockWidthC[], unsigned int ReadBlockHeightY[], unsigned int ReadBlockHeightC[], + unsigned int DCCMetaPitchY[], + unsigned int DCCMetaPitchC[], /* Output */ unsigned int SurfaceSizeInMALL[], @@ -1093,9 +1096,12 @@ void dml32_CalculatePrefetchBandwithSupport(unsigned int NumberOfActiveSurfaces, double UrgentBurstFactorLumaPre[], double UrgentBurstFactorChromaPre[], double UrgentBurstFactorCursorPre[], + double PrefetchBW[], + double VRatio[], + double MaxVRatioPre, /* output */ - double *PrefetchBandwidth, + double *MaxPrefetchBandwidth, double *FractionOfUrgentBandwidth, bool *PrefetchBandwidthSupport); @@ -1157,6 +1163,6 @@ bool dml32_CalculateDETSwathFillLatencyHiding(unsigned int NumberOfActiveSurface double PixelClock[], double VRatioY[], double VRatioC[], - enum dm_use_mall_for_pstate_change_mode UsesMALLForPStateChange[DC__NUM_DPP__MAX]); + enum dm_use_mall_for_pstate_change_mode UsesMALLForPStateChange[]); #endif diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn321/dcn321_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn321/dcn321_fpu.c index f4b176599be7..b80cef70fa60 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn321/dcn321_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn321/dcn321_fpu.c @@ -136,7 +136,7 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_21_soc = { .urgent_out_of_order_return_per_channel_pixel_only_bytes = 4096, .urgent_out_of_order_return_per_channel_pixel_and_vm_bytes = 4096, .urgent_out_of_order_return_per_channel_vm_only_bytes = 4096, - .pct_ideal_sdp_bw_after_urgent = 100.0, + .pct_ideal_sdp_bw_after_urgent = 90.0, .pct_ideal_fabric_bw_after_urgent = 67.0, .pct_ideal_dram_sdp_bw_after_urgent_pixel_only = 20.0, .pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm = 60.0, // N/A, for now keep as is until DML implemented @@ -534,8 +534,11 @@ void dcn321_update_bw_bounding_box_fpu(struct dc *dc, struct clk_bw_params *bw_p } /* Override from VBIOS for num_chan */ - if (dc->ctx->dc_bios->vram_info.num_chans) + if (dc->ctx->dc_bios->vram_info.num_chans) { dcn3_21_soc.num_chans = dc->ctx->dc_bios->vram_info.num_chans; + dcn3_21_soc.mall_allocated_for_dcn_mbytes = (double)(dcn32_calc_num_avail_chans_for_mall(dc, + dc->ctx->dc_bios->vram_info.num_chans) * dc->caps.mall_size_per_mem_channel); + } if (dc->ctx->dc_bios->vram_info.dram_channel_width_bytes) dcn3_21_soc.dram_channel_width_bytes = dc->ctx->dc_bios->vram_info.dram_channel_width_bytes; diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h index 3d643d50c3eb..a9d49ef58fb5 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h @@ -91,6 +91,7 @@ struct display_mode_lib { struct dal_logger *logger; struct dml_funcs funcs; struct _vcs_dpi_display_e2e_pipe_params_st dml_pipe_state[6]; + bool validate_max_state; }; void dml_init_instance(struct display_mode_lib *lib, diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h index 64d602e6412f..3c077164f362 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h @@ -246,6 +246,7 @@ struct _vcs_dpi_soc_bounding_box_st { bool disable_dram_clock_change_vactive_support; bool allow_dram_clock_one_display_vactive; enum self_refresh_affinity allow_dram_self_refresh_or_dram_clock_change_in_vblank; + double max_vratio_pre; }; /** diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c index 8e6585dab20e..f9653f511baa 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c @@ -202,6 +202,7 @@ dml_get_pipe_attr_func(vm_group_size_in_bytes, mode_lib->vba.vm_group_bytes); dml_get_pipe_attr_func(dpte_row_height_linear_l, mode_lib->vba.dpte_row_height_linear); dml_get_pipe_attr_func(pte_buffer_mode, mode_lib->vba.PTE_BUFFER_MODE); dml_get_pipe_attr_func(subviewport_lines_needed_in_mall, mode_lib->vba.SubViewportLinesNeededInMALL); +dml_get_pipe_attr_func(surface_size_in_mall, mode_lib->vba.SurfaceSizeInMALL) double get_total_immediate_flip_bytes( struct display_mode_lib *mode_lib, @@ -411,6 +412,7 @@ static void fetch_socbb_params(struct display_mode_lib *mode_lib) soc->urgent_latency_adjustment_fabric_clock_component_us; mode_lib->vba.UrgentLatencyAdjustmentFabricClockReference = soc->urgent_latency_adjustment_fabric_clock_reference_mhz; + mode_lib->vba.MaxVRatioPre = soc->max_vratio_pre; } static void fetch_ip_params(struct display_mode_lib *mode_lib) diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h index 81e53e67cd0b..07993741f5e6 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h @@ -143,6 +143,7 @@ dml_get_pipe_attr_decl(vready_at_or_after_vsync); dml_get_pipe_attr_decl(min_dst_y_next_start); dml_get_pipe_attr_decl(vstartup_calculated); dml_get_pipe_attr_decl(subviewport_lines_needed_in_mall); +dml_get_pipe_attr_decl(surface_size_in_mall); double get_total_immediate_flip_bytes( struct display_mode_lib *mode_lib, @@ -262,6 +263,7 @@ struct vba_vars_st { int maxMpcComb; bool UseMaximumVStartup; + double MaxVRatioPre; double WritebackDISPCLK; double DPPCLKUsingSingleDPPLuma; double DPPCLKUsingSingleDPPChroma; |