Merge tag 'amd-drm-fixes-5.14-2021-08-18' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-5.14-2021-08-18:

amdgpu:
- vega10 SMU workload fix
- DCN VM fix
- DCN 3.01 watermark fix

amdkfd:
- SVM fix

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210818225137.4070-1-alexander.deucher@amd.com

11 files changed, 53 insertions, 96 deletions

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
index c7b364e4a287..e883731c3f8f 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
@@ -3026,6 +3026,14 @@ svm_range_get_attr(struct kfd_process *p, uint64_t start, uint64_t size,
 	pr_debug("svms 0x%p [0x%llx 0x%llx] nattr 0x%x\n", &p->svms, start,
 		 start + size - 1, nattr);
 
+	/* Flush pending deferred work to avoid racing with deferred actions from
+	 * previous memory map changes (e.g. munmap). Concurrent memory map changes
+	 * can still race with get_attr because we don't hold the mmap lock. But that
+	 * would be a race condition in the application anyway, and undefined
+	 * behaviour is acceptable in that case.
+	 */
+	flush_work(&p->svms.deferred_list_work);
+
 	mmap_read_lock(mm);
 	if (!svm_range_is_valid(mm, start, size)) {
 		pr_debug("invalid range\n");
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 605e297b7a59..a30283fa5173 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -1530,6 +1530,12 @@ void dc_z10_restore(struct dc *dc)
 	if (dc->hwss.z10_restore)
 		dc->hwss.z10_restore(dc);
 }
+
+void dc_z10_save_init(struct dc *dc)
+{
+	if (dc->hwss.z10_save_init)
+		dc->hwss.z10_save_init(dc);
+}
 #endif
 /*
  * Applies given context to HW and copy it into current context.
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_vm_helper.c b/drivers/gpu/drm/amd/display/dc/core/dc_vm_helper.c
index f2b39ec35c89..cde8ed2560b3 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_vm_helper.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_vm_helper.c
@@ -47,6 +47,9 @@ int dc_setup_system_context(struct dc *dc, struct dc_phy_addr_space_config *pa_c
 		 */
 		memcpy(&dc->vm_pa_config, pa_config, sizeof(struct dc_phy_addr_space_config));
 		dc->vm_pa_config.valid = true;
+#if defined(CONFIG_DRM_AMD_DC_DCN)
+		dc_z10_save_init(dc);
+#endif
 	}
 
 	return num_vmids;
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index af7b60108e9d..21d78289b048 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -1338,6 +1338,7 @@ void dc_hardware_release(struct dc *dc);
 bool dc_set_psr_allow_active(struct dc *dc, bool enable);
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 void dc_z10_restore(struct dc *dc);
+void dc_z10_save_init(struct dc *dc);
 #endif
 
 bool dc_enable_dmub_notifications(struct dc *dc);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.c b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.c
index 9776d1737818..912285fdce18 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_resource.c
@@ -1622,106 +1622,12 @@ static void dcn301_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *b
 	dml_init_instance(&dc->dml, &dcn3_01_soc, &dcn3_01_ip, DML_PROJECT_DCN30);
 }
 
-static void calculate_wm_set_for_vlevel(
-		int vlevel,
-		struct wm_range_table_entry *table_entry,
-		struct dcn_watermarks *wm_set,
-		struct display_mode_lib *dml,
-		display_e2e_pipe_params_st *pipes,
-		int pipe_cnt)
-{
-	double dram_clock_change_latency_cached = dml->soc.dram_clock_change_latency_us;
-
-	ASSERT(vlevel < dml->soc.num_states);
-	/* only pipe 0 is read for voltage and dcf/soc clocks */
-	pipes[0].clks_cfg.voltage = vlevel;
-	pipes[0].clks_cfg.dcfclk_mhz = dml->soc.clock_limits[vlevel].dcfclk_mhz;
-	pipes[0].clks_cfg.socclk_mhz = dml->soc.clock_limits[vlevel].socclk_mhz;
-
-	dml->soc.dram_clock_change_latency_us = table_entry->pstate_latency_us;
-	dml->soc.sr_exit_time_us = table_entry->sr_exit_time_us;
-	dml->soc.sr_enter_plus_exit_time_us = table_entry->sr_enter_plus_exit_time_us;
-
-	wm_set->urgent_ns = get_wm_urgent(dml, pipes, pipe_cnt) * 1000;
-	wm_set->cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(dml, pipes, pipe_cnt) * 1000;
-	wm_set->cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(dml, pipes, pipe_cnt) * 1000;
-	wm_set->cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(dml, pipes, pipe_cnt) * 1000;
-	wm_set->pte_meta_urgent_ns = get_wm_memory_trip(dml, pipes, pipe_cnt) * 1000;
-	wm_set->frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(dml, pipes, pipe_cnt) * 1000;
-	wm_set->frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(dml, pipes, pipe_cnt) * 1000;
-	wm_set->urgent_latency_ns = get_urgent_latency(dml, pipes, pipe_cnt) * 1000;
-	dml->soc.dram_clock_change_latency_us = dram_clock_change_latency_cached;
-
-}
-
-static void dcn301_calculate_wm_and_dlg(
-		struct dc *dc, struct dc_state *context,
-		display_e2e_pipe_params_st *pipes,
-		int pipe_cnt,
-		int vlevel_req)
-{
-	int i, pipe_idx;
-	int vlevel, vlevel_max;
-	struct wm_range_table_entry *table_entry;
-	struct clk_bw_params *bw_params = dc->clk_mgr->bw_params;
-
-	ASSERT(bw_params);
-
-	vlevel_max = bw_params->clk_table.num_entries - 1;
-
-	/* WM Set D */
-	table_entry = &bw_params->wm_table.entries[WM_D];
-	if (table_entry->wm_type == WM_TYPE_RETRAINING)
-		vlevel = 0;
-	else
-		vlevel = vlevel_max;
-	calculate_wm_set_for_vlevel(vlevel, table_entry, &context->bw_ctx.bw.dcn.watermarks.d,
-						&context->bw_ctx.dml, pipes, pipe_cnt);
-	/* WM Set C */
-	table_entry = &bw_params->wm_table.entries[WM_C];
-	vlevel = min(max(vlevel_req, 2), vlevel_max);
-	calculate_wm_set_for_vlevel(vlevel, table_entry, &context->bw_ctx.bw.dcn.watermarks.c,
-						&context->bw_ctx.dml, pipes, pipe_cnt);
-	/* WM Set B */
-	table_entry = &bw_params->wm_table.entries[WM_B];
-	vlevel = min(max(vlevel_req, 1), vlevel_max);
-	calculate_wm_set_for_vlevel(vlevel, table_entry, &context->bw_ctx.bw.dcn.watermarks.b,
-						&context->bw_ctx.dml, pipes, pipe_cnt);
-
-	/* WM Set A */
-	table_entry = &bw_params->wm_table.entries[WM_A];
-	vlevel = min(vlevel_req, vlevel_max);
-	calculate_wm_set_for_vlevel(vlevel, table_entry, &context->bw_ctx.bw.dcn.watermarks.a,
-						&context->bw_ctx.dml, pipes, pipe_cnt);
-
-	for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) {
-		if (!context->res_ctx.pipe_ctx[i].stream)
-			continue;
-
-		pipes[pipe_idx].clks_cfg.dispclk_mhz = get_dispclk_calculated(&context->bw_ctx.dml, pipes, pipe_cnt);
-		pipes[pipe_idx].clks_cfg.dppclk_mhz = get_dppclk_calculated(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx);
-
-		if (dc->config.forced_clocks) {
-			pipes[pipe_idx].clks_cfg.dispclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dispclk_mhz;
-			pipes[pipe_idx].clks_cfg.dppclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dppclk_mhz;
-		}
-		if (dc->debug.min_disp_clk_khz > pipes[pipe_idx].clks_cfg.dispclk_mhz * 1000)
-			pipes[pipe_idx].clks_cfg.dispclk_mhz = dc->debug.min_disp_clk_khz / 1000.0;
-		if (dc->debug.min_dpp_clk_khz > pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000)
-			pipes[pipe_idx].clks_cfg.dppclk_mhz = dc->debug.min_dpp_clk_khz / 1000.0;
-
-		pipe_idx++;
-	}
-
-	dcn20_calculate_dlg_params(dc, context, pipes, pipe_cnt, vlevel);
-}
-
 static struct resource_funcs dcn301_res_pool_funcs = {
 	.destroy = dcn301_destroy_resource_pool,
 	.link_enc_create = dcn301_link_encoder_create,
 	.panel_cntl_create = dcn301_panel_cntl_create,
 	.validate_bandwidth = dcn30_validate_bandwidth,
-	.calculate_wm_and_dlg = dcn301_calculate_wm_and_dlg,
+	.calculate_wm_and_dlg = dcn30_calculate_wm_and_dlg,
 	.update_soc_for_wm_a = dcn30_update_soc_for_wm_a,
 	.populate_dml_pipes = dcn30_populate_dml_pipes_from_context,
 	.acquire_idle_pipe_for_layer = dcn20_acquire_idle_pipe_for_layer,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c
index 6ac6faf0c533..8a2119d8ca0d 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c
@@ -404,6 +404,18 @@ void dcn31_update_info_frame(struct pipe_ctx *pipe_ctx)
 			&pipe_ctx->stream_res.encoder_info_frame);
 	}
 }
+void dcn31_z10_save_init(struct dc *dc)
+{
+	union dmub_rb_cmd cmd;
+
+	memset(&cmd, 0, sizeof(cmd));
+	cmd.dcn_restore.header.type = DMUB_CMD__IDLE_OPT;
+	cmd.dcn_restore.header.sub_type = DMUB_CMD__IDLE_OPT_DCN_SAVE_INIT;
+
+	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);
+}
 
 void dcn31_z10_restore(struct dc *dc)
 {
diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.h b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.h
index 40dfebe78fdd..140435e4f7ff 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.h
@@ -44,6 +44,7 @@ void dcn31_enable_power_gating_plane(
 void dcn31_update_info_frame(struct pipe_ctx *pipe_ctx);
 
 void dcn31_z10_restore(struct dc *dc);
+void dcn31_z10_save_init(struct dc *dc);
 
 void dcn31_hubp_pg_control(struct dce_hwseq *hws, unsigned int hubp_inst, bool power_on);
 int dcn31_init_sys_ctx(struct dce_hwseq *hws, struct dc *dc, struct dc_phy_addr_space_config *pa_config);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_init.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_init.c
index aaf2dbd095fe..b30d923471cb 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_init.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_init.c
@@ -97,6 +97,7 @@ static const struct hw_sequencer_funcs dcn31_funcs = {
 	.set_abm_immediate_disable = dcn21_set_abm_immediate_disable,
 	.set_pipe = dcn21_set_pipe,
 	.z10_restore = dcn31_z10_restore,
+	.z10_save_init = dcn31_z10_save_init,
 	.is_abm_supported = dcn31_is_abm_supported,
 	.set_disp_pattern_generator = dcn30_set_disp_pattern_generator,
 	.update_visual_confirm_color = dcn20_update_visual_confirm_color,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
index 5ab008e62b82..ad5f2adcc40d 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
@@ -237,6 +237,7 @@ struct hw_sequencer_funcs {
 			int width, int height, int offset);
 
 	void (*z10_restore)(struct dc *dc);
+	void (*z10_save_init)(struct dc *dc);
 
 	void (*update_visual_confirm_color)(struct dc *dc,
 			struct pipe_ctx *pipe_ctx,
diff --git a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
index 7c4734f905d9..7fafb8d6c1da 100644
--- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
+++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
@@ -856,6 +856,11 @@ enum dmub_cmd_idle_opt_type {
 	 * DCN hardware restore.
 	 */
 	DMUB_CMD__IDLE_OPT_DCN_RESTORE = 0,
+
+	/**
+	 * DCN hardware save.
+	 */
+	DMUB_CMD__IDLE_OPT_DCN_SAVE_INIT = 1
 };
 
 /**
diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c
index 25979106fd25..02e8c6e5448d 100644
--- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c
@@ -5127,6 +5127,13 @@ static int vega10_get_power_profile_mode(struct pp_hwmgr *hwmgr, char *buf)
 	return size;
 }
 
+static bool vega10_get_power_profile_mode_quirks(struct pp_hwmgr *hwmgr)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+
+	return (adev->pdev->device == 0x6860);
+}
+
 static int vega10_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, uint32_t size)
 {
 	struct vega10_hwmgr *data = hwmgr->backend;
@@ -5163,9 +5170,15 @@ static int vega10_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, ui
 	}
 
 out:
-	smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetWorkloadMask,
+	if (vega10_get_power_profile_mode_quirks(hwmgr))
+		smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetWorkloadMask,
+						1 << power_profile_mode,
+						NULL);
+	else
+		smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetWorkloadMask,
 						(!power_profile_mode) ? 0 : 1 << (power_profile_mode - 1),
 						NULL);
+
 	hwmgr->power_profile_mode = power_profile_mode;
 
 	return 0;