drm/amdgpu/display: add MALL support (v2)

Enable Memory Access at Last Level (MALL) feature for display.

v2: squash in 64 bit division fixes

Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

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

drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c

@@ -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)

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);

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);
 

drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c

@@ -696,6 +696,10 @@ void dcn30_program_dmdata_engine(struct pipe_ctx *pipe_ctx)
 
 bool dcn30_apply_idle_power_optimizations(struct dc *dc, bool enable)
 {
+	union dmub_rb_cmd cmd;
+	unsigned int surface_size, refresh_hz, denom;
+	uint32_t tmr_delay = 0, tmr_scale = 0;
+
 	if (!dc->ctx->dmub_srv)
 		return false;
 
@@ -710,12 +714,75 @@ bool dcn30_apply_idle_power_optimizations(struct dc *dc, bool enable)
 					/* Fail eligibility on a visible stream */
 					break;
 			}
+
+			// TODO: remove hard code size
+			if (surface_size < 128 * 1024 * 1024) {
+				refresh_hz = div_u64((unsigned long long) dc->current_state->streams[0]->timing.pix_clk_100hz *
+						     100LL,
+						     (dc->current_state->streams[0]->timing.v_total *
+						      dc->current_state->streams[0]->timing.h_total));
+
+				/*
+				 * Delay_Us = 65.28 * (64 + MallFrameCacheTmrDly) * 2^MallFrameCacheTmrScale
+				 * Delay_Us / 65.28 = (64 + MallFrameCacheTmrDly) * 2^MallFrameCacheTmrScale
+				 * (Delay_Us / 65.28) / 2^MallFrameCacheTmrScale = 64 + MallFrameCacheTmrDly
+				 * MallFrameCacheTmrDly = ((Delay_Us / 65.28) / 2^MallFrameCacheTmrScale) - 64
+				 *                      = (1000000 / refresh) / 65.28 / 2^MallFrameCacheTmrScale - 64
+				 *                      = 1000000 / (refresh * 65.28 * 2^MallFrameCacheTmrScale) - 64
+				 *                      = (1000000 * 100) / (refresh * 6528 * 2^MallFrameCacheTmrScale) - 64
+				 *
+				 * need to round up the result of the division before the subtraction
+				 */
+				denom = refresh_hz * 6528;
+				tmr_delay = div_u64((100000000LL + denom - 1), denom) - 64LL;
+
+				/* scale should be increased until it fits into 6 bits */
+				while (tmr_delay & ~0x3F) {
+					tmr_scale++;
+
+					if (tmr_scale > 3) {
+						/* The delay exceeds the range of the hystersis timer */
+						ASSERT(false);
+						return false;
+					}
+
+					denom *= 2;
+					tmr_delay = div_u64((100000000LL + denom - 1), denom) - 64LL;
+				}
+
+				/* Enable MALL */
+				memset(&cmd, 0, sizeof(cmd));
+				cmd.mall.header.type = DMUB_CMD__MALL;
+				cmd.mall.header.sub_type =
+					DMUB_CMD__MALL_ACTION_ALLOW;
+				cmd.mall.header.payload_bytes =
+					sizeof(cmd.mall) -
+					sizeof(cmd.mall.header);
+				cmd.mall.tmr_delay = tmr_delay;
+				cmd.mall.tmr_scale = tmr_scale;
+
+				dc_dmub_srv_cmd_queue(dc->ctx->dmub_srv, &cmd);
+				dc_dmub_srv_cmd_execute(dc->ctx->dmub_srv);
+
+				return true;
+			}
 		}
 
 		/* No applicable optimizations */
 		return false;
 	}
 
+	/* Disable MALL */
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.mall.header.type = DMUB_CMD__MALL;
+	cmd.mall.header.sub_type = DMUB_CMD__MALL_ACTION_DISALLOW;
+	cmd.mall.header.payload_bytes =
+		sizeof(cmd.mall) - sizeof(cmd.mall.header);
+
+	dc_dmub_srv_cmd_queue(dc->ctx->dmub_srv, &cmd);
+	dc_dmub_srv_cmd_execute(dc->ctx->dmub_srv);
+	dc_dmub_srv_wait_idle(dc->ctx->dmub_srv);
+
 	return true;
 }
 

drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c

@@ -2247,7 +2247,7 @@ void dcn30_calculate_wm_and_dlg(
 	/* Set D:
 	 * DCFCLK: Min Required
 	 * FCLK(proportional to UCLK): 1GHz or Max
-	 * sr_enter_exit = 4, sr_exit = 2us
+	 * MALL stutter, sr_enter_exit = 4, sr_exit = 2us
 	 */
 	/*
 	if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].valid) {

drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h

@@ -298,6 +298,7 @@ enum dmub_cmd_type {
 	DMUB_CMD__REG_REG_WAIT = 4,
 	DMUB_CMD__PLAT_54186_WA = 5,
 	DMUB_CMD__PSR = 64,
+	DMUB_CMD__MALL = 65,
 	DMUB_CMD__ABM = 66,
 	DMUB_CMD__HW_LOCK = 69,
 	DMUB_CMD__DP_AUX_ACCESS = 70,
@@ -425,6 +426,18 @@ struct dmub_rb_cmd_PLAT_54186_wa {
 	struct dmub_cmd_PLAT_54186_wa flip;
 };
 
+struct dmub_rb_cmd_mall {
+	struct dmub_cmd_header header;
+	union dmub_addr cursor_copy_src;
+	union dmub_addr cursor_copy_dst;
+	uint32_t tmr_delay;
+	uint32_t tmr_scale;
+	uint16_t cursor_width;
+	uint16_t cursor_pitch;
+	uint16_t cursor_height;
+	uint8_t cursor_bpp;
+};
+
 struct dmub_cmd_digx_encoder_control_data {
 	union dig_encoder_control_parameters_v1_5 dig;
 };
@@ -556,6 +569,12 @@ enum psr_version {
 	PSR_VERSION_UNSUPPORTED			= 0xFFFFFFFF,
 };
 
+enum dmub_cmd_mall_type {
+	DMUB_CMD__MALL_ACTION_ALLOW = 0,
+	DMUB_CMD__MALL_ACTION_DISALLOW = 1,
+	DMUB_CMD__MALL_ACTION_COPY_CURSOR = 2,
+};
+
 struct dmub_cmd_psr_copy_settings_data {
 	union dmub_psr_debug_flags debug;
 	uint16_t psr_level;
@@ -761,6 +780,7 @@ union dmub_rb_cmd {
 	struct dmub_rb_cmd_psr_enable psr_enable;
 	struct dmub_rb_cmd_psr_set_level psr_set_level;
 	struct dmub_rb_cmd_PLAT_54186_wa PLAT_54186_wa;
+	struct dmub_rb_cmd_mall mall;
 	struct dmub_rb_cmd_abm_set_pipe abm_set_pipe;
 	struct dmub_rb_cmd_abm_set_backlight abm_set_backlight;
 	struct dmub_rb_cmd_abm_set_level abm_set_level;