From b0814fa3be76a8c62cbb9e02bb851b0ec234037d Mon Sep 17 00:00:00 2001 From: Alex Hung Date: Wed, 16 Oct 2024 12:18:39 -0600 Subject: [PATCH 01/16] drm/amd/display: Remove useless assignments and variables [WHAT & HOW] misc0, temp and split_pipe are assigned but immediately re-assigned to other values. The early assignments are useless and are removed. Unused variables are removed as well. This fixes 5 UNUSED_VALUE issues reported by Coverity. Reviewed-by: Rodrigo Siqueira Signed-off-by: Alex Hung Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c | 1 - drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c | 2 -- drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c | 2 -- drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c | 2 -- .../drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c | 2 -- 5 files changed, 9 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c index f0417ee6fcf8..f90fc154549a 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c +++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c @@ -1013,7 +1013,6 @@ static bool dc_can_pipe_disable_cursor(struct pipe_ctx *pipe_ctx) r2 = test_pipe->plane_res.scl_data.recout; r2_r = r2.x + r2.width; r2_b = r2.y + r2.height; - split_pipe = test_pipe; /** * There is another half plane on same layer because of diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c index 5c2825bc9a87..d199e4ed2e59 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c @@ -277,7 +277,6 @@ static void dce110_stream_encoder_dp_set_stream_attribute( uint32_t misc1 = 0; uint32_t h_blank; uint32_t h_back_porch; - uint8_t synchronous_clock = 0; /* asynchronous mode */ uint8_t colorimetry_bpc; uint8_t dynamic_range_rgb = 0; /*full range*/ uint8_t dynamic_range_ycbcr = 1; /*bt709*/ @@ -380,7 +379,6 @@ static void dce110_stream_encoder_dp_set_stream_attribute( break; } - misc0 = misc0 | synchronous_clock; misc0 = colorimetry_bpc << 5; if (REG(DP_MSA_TIMING_PARAM1)) { diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c index db7557a1c613..8a3fbf95c48f 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c @@ -76,7 +76,6 @@ UNP_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH_C__GRPH_PRIMARY_SURFACE_ADDRESS_HIGH_C_MAS mmUNP_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH_C, value); - temp = 0; value = 0; temp = address.low_part >> UNP_GRPH_PRIMARY_SURFACE_ADDRESS_C__GRPH_PRIMARY_SURFACE_ADDRESS_C__SHIFT; @@ -112,7 +111,6 @@ UNP_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH_L__GRPH_PRIMARY_SURFACE_ADDRESS_HIGH_L_MAS mmUNP_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH_L, value); - temp = 0; value = 0; temp = address.low_part >> UNP_GRPH_PRIMARY_SURFACE_ADDRESS_L__GRPH_PRIMARY_SURFACE_ADDRESS_L__SHIFT; diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c index f496e952ceec..d01a8b8f9595 100644 --- a/drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c @@ -255,7 +255,6 @@ void enc1_stream_encoder_dp_set_stream_attribute( uint32_t misc1 = 0; uint32_t h_blank; uint32_t h_back_porch; - uint8_t synchronous_clock = 0; /* asynchronous mode */ uint8_t colorimetry_bpc; uint8_t dp_pixel_encoding = 0; uint8_t dp_component_depth = 0; @@ -362,7 +361,6 @@ void enc1_stream_encoder_dp_set_stream_attribute( break; } - misc0 = misc0 | synchronous_clock; misc0 = colorimetry_bpc << 5; switch (output_color_space) { diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c index 0a27e0942a12..098c2a01a850 100644 --- a/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c +++ b/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c @@ -447,7 +447,6 @@ void enc401_stream_encoder_dp_set_stream_attribute( uint32_t misc1 = 0; uint32_t h_blank; uint32_t h_back_porch; - uint8_t synchronous_clock = 0; /* asynchronous mode */ uint8_t colorimetry_bpc; uint8_t dp_pixel_encoding = 0; uint8_t dp_component_depth = 0; @@ -603,7 +602,6 @@ void enc401_stream_encoder_dp_set_stream_attribute( break; } - misc0 = misc0 | synchronous_clock; misc0 = colorimetry_bpc << 5; switch (output_color_space) { -- 2.51.0 From d2bf27be839e89c6fd24b3ad3a2b38dcbfbf378a Mon Sep 17 00:00:00 2001 From: Alex Hung Date: Wed, 16 Oct 2024 12:23:58 -0600 Subject: [PATCH 02/16] drm/amd/display: Simplify dcn35_is_ips_supported() [WHAT & HOW] The variable "ips_supported" is redundant and we can return from dcn35_smu_get_ips_supported directly. This fixes 1 UNUSED_VALUE issue reported by Coverity. Reviewed-by: Rodrigo Siqueira Signed-off-by: Alex Hung Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c index 90208a5096c0..1d4fe0de0f67 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c @@ -978,11 +978,8 @@ static void dcn35_exit_low_power_state(struct clk_mgr *clk_mgr_base) static bool dcn35_is_ips_supported(struct clk_mgr *clk_mgr_base) { struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); - bool ips_supported = true; - ips_supported = dcn35_smu_get_ips_supported(clk_mgr) ? true : false; - - return ips_supported; + return dcn35_smu_get_ips_supported(clk_mgr) ? true : false; } static void dcn35_init_clocks_fpga(struct clk_mgr *clk_mgr) -- 2.51.0 From 7ef6f3ae4cd21a4ab86e04c7f11a6bdd92332b60 Mon Sep 17 00:00:00 2001 From: Joshua Aberback Date: Thu, 17 Oct 2024 15:56:30 -0400 Subject: [PATCH 03/16] drm/amd/display: Change MPC Tree visual confirm colours [Why] MPC background colours that use fractional components look different if MPC OGAM is in use vs in bypass mode. The current red and orange colours look very similar when OGAM is in bypass, so the colours need to change to be consistently very easy to tell apart. [How] Use colours that only have 0 or MAX values in each component Reviewed-by: Alvin Lee Signed-off-by: Joshua Aberback Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c index 2fdcf8d59b9f..0419ee7f22a5 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c @@ -312,11 +312,11 @@ void get_mpctree_visual_confirm_color( { const struct tg_color pipe_colors[6] = { {MAX_TG_COLOR_VALUE, 0, 0}, /* red */ - {MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE / 4, 0}, /* orange */ {MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE, 0}, /* yellow */ {0, MAX_TG_COLOR_VALUE, 0}, /* green */ + {0, MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE}, /* cyan */ {0, 0, MAX_TG_COLOR_VALUE}, /* blue */ - {MAX_TG_COLOR_VALUE / 2, 0, MAX_TG_COLOR_VALUE / 2}, /* purple */ + {MAX_TG_COLOR_VALUE, 0, MAX_TG_COLOR_VALUE}, /* magenta */ }; struct pipe_ctx *top_pipe = pipe_ctx; -- 2.51.0 From 1b7ac448cc544f6a4f8543423d9c2b726f3313fd Mon Sep 17 00:00:00 2001 From: Aurabindo Pillai Date: Wed, 16 Oct 2024 13:08:02 -0400 Subject: [PATCH 04/16] drm/amd/display: Fix idle optimizations entry log [Why & How] Whether we really enter idle optimizations are decided within DC. Printing into dmesg before calling the DC API gives an incorrect indication that we are entering idle optimization in cases where its disabled manually. To fix this, remove the print in DM and add them in DC Reviewed-by: Alvin Lee Signed-off-by: Aurabindo Pillai Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c | 8 ++------ drivers/gpu/drm/amd/display/dc/core/dc.c | 8 ++++++-- 2 files changed, 8 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c index 171099a3ea05..8b5bea799a24 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c @@ -252,10 +252,8 @@ static void amdgpu_dm_crtc_vblank_control_worker(struct work_struct *work) else if (dm->active_vblank_irq_count) dm->active_vblank_irq_count--; - if (dm->active_vblank_irq_count > 0) { - DRM_DEBUG_KMS("Allow idle optimizations (MALL): false\n"); + if (dm->active_vblank_irq_count > 0) dc_allow_idle_optimizations(dm->dc, false); - } /* * Control PSR based on vblank requirements from OS @@ -274,10 +272,8 @@ static void amdgpu_dm_crtc_vblank_control_worker(struct work_struct *work) vblank_work->stream->link->replay_settings.replay_feature_enabled); } - if (dm->active_vblank_irq_count == 0) { - DRM_DEBUG_KMS("Allow idle optimizations (MALL): true\n"); + if (dm->active_vblank_irq_count == 0) dc_allow_idle_optimizations(dm->dc, true); - } mutex_unlock(&dm->dc_lock); diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index 5a12fc75f97f..8a52fef46785 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -5429,8 +5429,10 @@ bool dc_set_ips_disable(struct dc *dc, unsigned int disable_ips) void dc_allow_idle_optimizations_internal(struct dc *dc, bool allow, char const *caller_name) { - if (dc->debug.disable_idle_power_optimizations) + if (dc->debug.disable_idle_power_optimizations) { + DC_LOG_DEBUG("%s: disabled\n", __func__); return; + } if (allow != dc->idle_optimizations_allowed) DC_LOG_IPS("%s: allow_idle old=%d new=%d (caller=%s)\n", __func__, @@ -5447,8 +5449,10 @@ void dc_allow_idle_optimizations_internal(struct dc *dc, bool allow, char const return; if (dc->hwss.apply_idle_power_optimizations && dc->clk_mgr != NULL && - dc->hwss.apply_idle_power_optimizations(dc, allow)) + dc->hwss.apply_idle_power_optimizations(dc, allow)) { dc->idle_optimizations_allowed = allow; + DC_LOG_DEBUG("%s: %s\n", __func__, allow ? "enabled" : "disabled"); + } } void dc_exit_ips_for_hw_access_internal(struct dc *dc, const char *caller_name) -- 2.51.0 From 558cec793e73e5d22c96c56b1f70c83a8ce4b672 Mon Sep 17 00:00:00 2001 From: Ovidiu Bunea Date: Tue, 15 Oct 2024 18:20:54 -0400 Subject: [PATCH 05/16] drm/amd/display: Do not read DSC state if not in use [why & how] DSC may be power gated when coming out of S0i3, so avoid polling DSC registers since it will fail anyways. Only read if it is known that DSC is in use. Reviewed-by: Charlene Liu Signed-off-by: Ovidiu Bunea Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- .../drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c | 15 +++++++++------ 1 file changed, 9 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c index bd309dbdf7b2..3cb4e9907411 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c @@ -841,6 +841,7 @@ void dcn35_init_pipes(struct dc *dc, struct dc_state *context) uint32_t num_opps = 0; uint32_t opp_id_src0 = OPP_ID_INVALID; uint32_t opp_id_src1 = OPP_ID_INVALID; + uint32_t optc_dsc_state = 0; // Step 1: To find out which OPTC is running & OPTC DSC is ON // We can't use res_pool->res_cap->num_timing_generator to check @@ -849,7 +850,6 @@ void dcn35_init_pipes(struct dc *dc, struct dc_state *context) // Some ASICs would be fused display pipes less than the default setting. // In dcnxx_resource_construct function, driver would obatin real information. for (i = 0; i < dc->res_pool->timing_generator_count; i++) { - uint32_t optc_dsc_state = 0; struct timing_generator *tg = dc->res_pool->timing_generators[i]; if (tg->funcs->is_tg_enabled(tg)) { @@ -864,15 +864,18 @@ void dcn35_init_pipes(struct dc *dc, struct dc_state *context) } } - // Step 2: To power down DSC but skip DSC of running OPTC + // Step 2: To power down DSC but skip DSC of running OPTC for (i = 0; i < dc->res_pool->res_cap->num_dsc; i++) { struct dcn_dsc_state s = {0}; - dc->res_pool->dscs[i]->funcs->dsc_read_state(dc->res_pool->dscs[i], &s); + /* avoid reading DSC state when it is not in use as it may be power gated */ + if (optc_dsc_state) { + dc->res_pool->dscs[i]->funcs->dsc_read_state(dc->res_pool->dscs[i], &s); - if ((s.dsc_opp_source == opp_id_src0 || s.dsc_opp_source == opp_id_src1) && - s.dsc_clock_en && s.dsc_fw_en) - continue; + if ((s.dsc_opp_source == opp_id_src0 || s.dsc_opp_source == opp_id_src1) && + s.dsc_clock_en && s.dsc_fw_en) + continue; + } pg_cntl->funcs->dsc_pg_control(pg_cntl, dc->res_pool->dscs[i]->inst, false); } -- 2.51.0 From f3928f3d481920c748328192ec2ed4ab5d125d6b Mon Sep 17 00:00:00 2001 From: Samson Tam Date: Fri, 18 Oct 2024 00:26:41 -0400 Subject: [PATCH 06/16] drm/amd/display: store sharpness 1dlut table in dscl_prog_data [Why] Previously dscl_prog_data stored pointer to sharpness 1dlut table. SPL had four pre-generated tables, one for each setup. This allowed us to minimize number of times we had to recalculate table when switching between setups. However, with dual display, this becomes an issue because for a given setup, we could have a different per app sharpness value than the global sharpness value. So the pre-generated table will change but both displays may point to the same table and one of them will have the wrong sharpness setting. [How] Store the sharpness 1dlut table in dscl_prog_data. This ensures that each display can have its own sharpness setting. Reviewed-by: Ilya Bakoulin Signed-off-by: Samson Tam Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_dscl.c | 3 ++- drivers/gpu/drm/amd/display/dc/spl/dc_spl.c | 3 ++- drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.h | 1 - drivers/gpu/drm/amd/display/dc/spl/dc_spl_types.h | 3 ++- 4 files changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_dscl.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_dscl.c index 5105fd580017..2f92e7d4981b 100644 --- a/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_dscl.c +++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_dscl.c @@ -1091,7 +1091,8 @@ void dpp401_dscl_set_scaler_manual_scale(struct dpp *dpp_base, /* ISHARP_DELTA_LUT */ dpp401_dscl_set_isharp_filter(dpp, scl_data->dscl_prog_data.isharp_delta); dpp->scl_data.dscl_prog_data.sharpness_level = scl_data->dscl_prog_data.sharpness_level; - dpp->scl_data.dscl_prog_data.isharp_delta = scl_data->dscl_prog_data.isharp_delta; + memcpy(dpp->scl_data.dscl_prog_data.isharp_delta, scl_data->dscl_prog_data.isharp_delta, + sizeof(uint32_t) * ISHARP_LUT_TABLE_SIZE); if (memcmp(&dpp->scl_data, scl_data, sizeof(*scl_data)) == 0) return; diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl.c b/drivers/gpu/drm/amd/display/dc/spl/dc_spl.c index 403fd1221803..133906e73a65 100644 --- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl.c +++ b/drivers/gpu/drm/amd/display/dc/spl/dc_spl.c @@ -1607,7 +1607,8 @@ static void spl_set_isharp_data(struct dscl_prog_data *dscl_prog_data, spl_build_isharp_1dlut_from_reference_curve(ratio, setup, adp_sharpness, scale_to_sharpness_policy); - dscl_prog_data->isharp_delta = spl_get_pregen_filter_isharp_1D_lut(setup); + memcpy(dscl_prog_data->isharp_delta, spl_get_pregen_filter_isharp_1D_lut(setup), + sizeof(uint32_t) * ISHARP_LUT_TABLE_SIZE); dscl_prog_data->sharpness_level = adp_sharpness.sharpness_level; dscl_prog_data->isharp_en = 1; // ISHARP_EN diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.h b/drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.h index afcc66206ca2..89af91e19b6c 100644 --- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.h +++ b/drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.h @@ -7,7 +7,6 @@ #include "dc_spl_types.h" -#define ISHARP_LUT_TABLE_SIZE 32 const uint32_t *spl_get_filter_isharp_1D_lut_0(void); const uint32_t *spl_get_filter_isharp_1D_lut_0p5x(void); const uint32_t *spl_get_filter_isharp_1D_lut_1p0x(void); diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_types.h b/drivers/gpu/drm/amd/display/dc/spl/dc_spl_types.h index fcb5d389592b..8b00ccb1dfda 100644 --- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_types.h +++ b/drivers/gpu/drm/amd/display/dc/spl/dc_spl_types.h @@ -250,6 +250,7 @@ enum isharp_en { ISHARP_DISABLE, ISHARP_ENABLE }; +#define ISHARP_LUT_TABLE_SIZE 32 // Below struct holds values that can be directly used to program // hardware registers. No conversion/clamping is required struct dscl_prog_data { @@ -400,7 +401,7 @@ struct dscl_prog_data { uint32_t isharp_nl_en; // ISHARP_NL_EN ? TODO:check this struct isharp_lba isharp_lba; // ISHARP_LBA struct isharp_fmt isharp_fmt; // ISHARP_FMT - const uint32_t *isharp_delta; + uint32_t isharp_delta[ISHARP_LUT_TABLE_SIZE]; struct isharp_nldelta_sclip isharp_nldelta_sclip; // ISHARP_NLDELTA_SCLIP /* blur and scale filter */ const uint16_t *filter_blur_scale_v; -- 2.51.0 From 12cfb5d8eaefbb594dbb0a5a58874e8c5aefba13 Mon Sep 17 00:00:00 2001 From: Taimur Hassan Date: Sun, 20 Oct 2024 03:35:34 -0400 Subject: [PATCH 07/16] drm/amd/display: [FW Promotion] Release 0.0.240.0 Add some scruct for secure display. Acked-by: Wayne Lin Signed-off-by: Taimur Hassan Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/dmub/inc/dmub_cmd.h | 53 +++++++++++++++---- 1 file changed, 44 insertions(+), 9 deletions(-) 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 6d96a840d24d..3aa6c60588b5 100644 --- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h @@ -727,6 +727,7 @@ enum dmub_shared_state_feature_id { DMUB_SHARED_SHARE_FEATURE__INVALID = 0, DMUB_SHARED_SHARE_FEATURE__IPS_FW = 1, DMUB_SHARED_SHARE_FEATURE__IPS_DRIVER = 2, + DMUB_SHARED_SHARE_FEATURE__DEBUG_SETUP = 3, DMUB_SHARED_STATE_FEATURE__LAST, /* Total number of features. */ }; @@ -764,6 +765,14 @@ union dmub_shared_state_ips_driver_signals { */ #define DMUB_SHARED_STATE__IPS_FW_VERSION 1 +struct dmub_shared_state_debug_setup { + union { + struct { + uint32_t exclude_points[62]; + } profile_mode; + }; +}; + /** * struct dmub_shared_state_ips_fw - Firmware state for IPS. */ @@ -816,6 +825,7 @@ struct dmub_shared_state_feature_block { struct dmub_shared_state_feature_common common; /**< Generic data */ struct dmub_shared_state_ips_fw ips_fw; /**< IPS firmware state */ struct dmub_shared_state_ips_driver ips_driver; /**< IPS driver state */ + struct dmub_shared_state_debug_setup debug_setup; /**< Debug setup */ } data; /**< Shared state data. */ }; /* 256-bytes, fixed */ @@ -1158,6 +1168,10 @@ enum dmub_gpint_command { * RETURN: Total residency in microseconds - upper 32 bits */ DMUB_GPINT__GET_IPS_RESIDENCY_DURATION_US_HI = 133, + /** + * DESC: Setup debug configs. + */ + DMUB_GPINT__SETUP_DEBUG_MODE = 136, }; /** @@ -5172,7 +5186,34 @@ struct dmub_rb_cmd_get_usbc_cable_id { enum dmub_cmd_secure_display_type { DMUB_CMD__SECURE_DISPLAY_TEST_CMD = 0, /* test command to only check if inbox message works */ DMUB_CMD__SECURE_DISPLAY_CRC_STOP_UPDATE, - DMUB_CMD__SECURE_DISPLAY_CRC_WIN_NOTIFY + DMUB_CMD__SECURE_DISPLAY_CRC_WIN_NOTIFY, + DMUB_CMD__SECURE_DISPLAY_MULTIPLE_CRC_STOP_UPDATE, + DMUB_CMD__SECURE_DISPLAY_MULTIPLE_CRC_WIN_NOTIFY +}; + +#define MAX_ROI_NUM 2 + +struct dmub_cmd_roi_info { + uint16_t x_start; + uint16_t x_end; + uint16_t y_start; + uint16_t y_end; + uint8_t otg_id; + uint8_t phy_id; +}; + +struct dmub_cmd_roi_window_ctl { + uint16_t x_start; + uint16_t x_end; + uint16_t y_start; + uint16_t y_end; + bool enable; +}; + +struct dmub_cmd_roi_ctl_info { + uint8_t otg_id; + uint8_t phy_id; + struct dmub_cmd_roi_window_ctl roi_ctl[MAX_ROI_NUM]; }; /** @@ -5183,14 +5224,8 @@ struct dmub_rb_cmd_secure_display { /** * Data passed from driver to dmub firmware. */ - struct dmub_cmd_roi_info { - uint16_t x_start; - uint16_t x_end; - uint16_t y_start; - uint16_t y_end; - uint8_t otg_id; - uint8_t phy_id; - } roi_info; + struct dmub_cmd_roi_info roi_info; + struct dmub_cmd_roi_ctl_info mul_roi_ctl; }; /** -- 2.51.0 From cc1977d86e0109de03efe02682faf3775af56fb8 Mon Sep 17 00:00:00 2001 From: Aric Cyr Date: Sun, 20 Oct 2024 21:41:45 -0400 Subject: [PATCH 08/16] drm/amd/display: 3.2.307 This version brings along following fixes: - Fix polling DSC registers during S0i3 - Fix idle optimizations entry log - Change MPC Tree visual confirm colours - Fix underflow when playing 8K video in full screen mode - Optimize power up sequence for specific OLED Acked-by: Wayne Lin Signed-off-by: Aric Cyr Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dc.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 412cdb01a61a..72adbab589f5 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -55,7 +55,7 @@ struct aux_payload; struct set_config_cmd_payload; struct dmub_notification; -#define DC_VER "3.2.306" +#define DC_VER "3.2.307" #define MAX_SURFACES 3 #define MAX_PLANES 6 -- 2.51.0 From 0880f58f9609f0200483a49429af0f050d281703 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Fri, 25 Oct 2024 15:56:39 +0100 Subject: [PATCH 09/16] drm/amd/pm: Vangogh: Fix kernel memory out of bounds write KASAN reports that the GPU metrics table allocated in vangogh_tables_init() is not large enough for the memset done in smu_cmn_init_soft_gpu_metrics(). Condensed report follows: [ 33.861314] BUG: KASAN: slab-out-of-bounds in smu_cmn_init_soft_gpu_metrics+0x73/0x200 [amdgpu] [ 33.861799] Write of size 168 at addr ffff888129f59500 by task mangoapp/1067 ... [ 33.861808] CPU: 6 UID: 1000 PID: 1067 Comm: mangoapp Tainted: G W 6.12.0-rc4 #356 1a56f59a8b5182eeaf67eb7cb8b13594dd23b544 [ 33.861816] Tainted: [W]=WARN [ 33.861818] Hardware name: Valve Galileo/Galileo, BIOS F7G0107 12/01/2023 [ 33.861822] Call Trace: [ 33.861826] [ 33.861829] dump_stack_lvl+0x66/0x90 [ 33.861838] print_report+0xce/0x620 [ 33.861853] kasan_report+0xda/0x110 [ 33.862794] kasan_check_range+0xfd/0x1a0 [ 33.862799] __asan_memset+0x23/0x40 [ 33.862803] smu_cmn_init_soft_gpu_metrics+0x73/0x200 [amdgpu 13b1bc364ec578808f676eba412c20eaab792779] [ 33.863306] vangogh_get_gpu_metrics_v2_4+0x123/0xad0 [amdgpu 13b1bc364ec578808f676eba412c20eaab792779] [ 33.864257] vangogh_common_get_gpu_metrics+0xb0c/0xbc0 [amdgpu 13b1bc364ec578808f676eba412c20eaab792779] [ 33.865682] amdgpu_dpm_get_gpu_metrics+0xcc/0x110 [amdgpu 13b1bc364ec578808f676eba412c20eaab792779] [ 33.866160] amdgpu_get_gpu_metrics+0x154/0x2d0 [amdgpu 13b1bc364ec578808f676eba412c20eaab792779] [ 33.867135] dev_attr_show+0x43/0xc0 [ 33.867147] sysfs_kf_seq_show+0x1f1/0x3b0 [ 33.867155] seq_read_iter+0x3f8/0x1140 [ 33.867173] vfs_read+0x76c/0xc50 [ 33.867198] ksys_read+0xfb/0x1d0 [ 33.867214] do_syscall_64+0x90/0x160 ... [ 33.867353] Allocated by task 378 on cpu 7 at 22.794876s: [ 33.867358] kasan_save_stack+0x33/0x50 [ 33.867364] kasan_save_track+0x17/0x60 [ 33.867367] __kasan_kmalloc+0x87/0x90 [ 33.867371] vangogh_init_smc_tables+0x3f9/0x840 [amdgpu] [ 33.867835] smu_sw_init+0xa32/0x1850 [amdgpu] [ 33.868299] amdgpu_device_init+0x467b/0x8d90 [amdgpu] [ 33.868733] amdgpu_driver_load_kms+0x19/0xf0 [amdgpu] [ 33.869167] amdgpu_pci_probe+0x2d6/0xcd0 [amdgpu] [ 33.869608] local_pci_probe+0xda/0x180 [ 33.869614] pci_device_probe+0x43f/0x6b0 Empirically we can confirm that the former allocates 152 bytes for the table, while the latter memsets the 168 large block. Root cause appears that when GPU metrics tables for v2_4 parts were added it was not considered to enlarge the table to fit. The fix in this patch is rather "brute force" and perhaps later should be done in a smarter way, by extracting and consolidating the part version to size logic to a common helper, instead of brute forcing the largest possible allocation. Nevertheless, for now this works and fixes the out of bounds write. v2: * Drop impossible v3_0 case. (Mario) Signed-off-by: Tvrtko Ursulin Fixes: 41cec40bc9ba ("drm/amd/pm: Vangogh: Add new gpu_metrics_v2_4 to acquire gpu_metrics") Cc: Mario Limonciello Cc: Evan Quan Cc: Wenyou Yang Cc: Alex Deucher Reviewed-by: Mario Limonciello Link: https://lore.kernel.org/r/20241025145639.19124-1-tursulin@igalia.com Signed-off-by: Mario Limonciello Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c index a333ab827f48..6c43724c01dd 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c @@ -242,7 +242,9 @@ static int vangogh_tables_init(struct smu_context *smu) goto err0_out; smu_table->metrics_time = 0; - smu_table->gpu_metrics_table_size = max(sizeof(struct gpu_metrics_v2_3), sizeof(struct gpu_metrics_v2_2)); + smu_table->gpu_metrics_table_size = sizeof(struct gpu_metrics_v2_2); + smu_table->gpu_metrics_table_size = max(smu_table->gpu_metrics_table_size, sizeof(struct gpu_metrics_v2_3)); + smu_table->gpu_metrics_table_size = max(smu_table->gpu_metrics_table_size, sizeof(struct gpu_metrics_v2_4)); smu_table->gpu_metrics_table = kzalloc(smu_table->gpu_metrics_table_size, GFP_KERNEL); if (!smu_table->gpu_metrics_table) goto err1_out; -- 2.51.0 From 58a8c756fc4ca243fb5c070e1b9e0970f00757d9 Mon Sep 17 00:00:00 2001 From: Prike Liang Date: Fri, 27 Sep 2024 16:05:21 +0800 Subject: [PATCH 10/16] drm/amdgpu: correct the S3 abort check condition In the normal S3 entry, the TOS cycle counter is not reset during BIOS execution the _S3 method, so it doesn't determine whether the _S3 method is executed exactly. Howerver, the PM core performs the S3 suspend will set the PM_SUSPEND_FLAG_FW_RESUME bit if all the devices suspend successfully. Therefore, drivers can check the pm_suspend_global_flags bit(1) to detect the S3 suspend abort event. Fixes: 6704dbf71928 ("drm/amdgpu: update suspend status for aborting from deeper suspend") Signed-off-by: Prike Liang Reviewed-by: Lijo Lazar Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/soc15.c | 23 +++++++++++++---------- 1 file changed, 13 insertions(+), 10 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index 93e44e7ee3fa..af739b13b6b5 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -578,16 +578,13 @@ soc15_asic_reset_method(struct amdgpu_device *adev) static bool soc15_need_reset_on_resume(struct amdgpu_device *adev) { - u32 sol_reg; - - sol_reg = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81); - /* Will reset for the following suspend abort cases. - * 1) Only reset limit on APU side, dGPU hasn't checked yet. - * 2) S3 suspend abort and TOS already launched. + * 1) Only reset on APU side, dGPU hasn't checked yet. + * 2) S3 suspend aborted in the normal S3 suspend or + * performing pm core test. */ if (adev->flags & AMD_IS_APU && adev->in_s3 && - sol_reg) { + !pm_resume_via_firmware()) { adev->suspend_complete = false; return true; } else { @@ -603,11 +600,17 @@ static int soc15_asic_reset(struct amdgpu_device *adev) * successfully. So now, temporarily enable it for the * S3 suspend abort case. */ - if (((adev->apu_flags & AMD_APU_IS_RAVEN) || - (adev->apu_flags & AMD_APU_IS_RAVEN2)) && - !soc15_need_reset_on_resume(adev)) + + if ((adev->apu_flags & AMD_APU_IS_PICASSO || + !(adev->apu_flags & AMD_APU_IS_RAVEN)) && + soc15_need_reset_on_resume(adev)) + goto asic_reset; + + if ((adev->apu_flags & AMD_APU_IS_RAVEN) || + (adev->apu_flags & AMD_APU_IS_RAVEN2)) return 0; +asic_reset: switch (soc15_asic_reset_method(adev)) { case AMD_RESET_METHOD_PCI: dev_info(adev->dev, "PCI reset\n"); -- 2.51.0 From d5e3d8a2a6cb8b8c8678e60ae8067c18ffbc2da2 Mon Sep 17 00:00:00 2001 From: Prike Liang Date: Mon, 14 Oct 2024 15:25:35 +0800 Subject: [PATCH 11/16] drm/amdgpu: clean up the suspend_complete To check the status of S3 suspend completion, use the PM core pm_suspend_global_flags bit(1) to detect S3 abort events. Therefore, clean up the AMDGPU driver's private flag suspend_complete. Signed-off-by: Prike Liang Reviewed-by: Lijo Lazar Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 2 -- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 2 -- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 4 ++-- drivers/gpu/drm/amd/amdgpu/soc15.c | 7 ++----- drivers/gpu/drm/amd/amdgpu/soc21.c | 5 +++-- 5 files changed, 7 insertions(+), 13 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 3af5acff8518..607998d8b1d5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -1111,8 +1111,6 @@ struct amdgpu_device { bool in_s3; bool in_s4; bool in_s0ix; - /* indicate amdgpu suspension status */ - bool suspend_complete; enum pp_mp1_state mp1_state; struct amdgpu_doorbell_index doorbell_index; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 46f756513948..ff3ac30744dc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -2500,7 +2500,6 @@ static int amdgpu_pmops_suspend(struct device *dev) struct drm_device *drm_dev = dev_get_drvdata(dev); struct amdgpu_device *adev = drm_to_adev(drm_dev); - adev->suspend_complete = false; if (amdgpu_acpi_is_s0ix_active(adev)) adev->in_s0ix = true; else if (amdgpu_acpi_is_s3_active(adev)) @@ -2515,7 +2514,6 @@ static int amdgpu_pmops_suspend_noirq(struct device *dev) struct drm_device *drm_dev = dev_get_drvdata(dev); struct amdgpu_device *adev = drm_to_adev(drm_dev); - adev->suspend_complete = true; if (amdgpu_acpi_should_gpu_reset(adev)) return amdgpu_asic_reset(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 66947850d7e4..e9248a855ba7 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -3288,8 +3288,8 @@ static int gfx_v9_0_cp_gfx_start(struct amdgpu_device *adev) * confirmed that the APU gfx10/gfx11 needn't such update. */ if (adev->flags & AMD_IS_APU && - adev->in_s3 && !adev->suspend_complete) { - DRM_INFO(" Will skip the CSB packet resubmit\n"); + adev->in_s3 && !pm_resume_via_firmware()) { + DRM_INFO("Will skip the CSB packet resubmit\n"); return 0; } r = amdgpu_ring_alloc(ring, gfx_v9_0_get_csb_size(adev) + 4 + 3); diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index af739b13b6b5..533b4b2b432d 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -584,13 +584,10 @@ static bool soc15_need_reset_on_resume(struct amdgpu_device *adev) * performing pm core test. */ if (adev->flags & AMD_IS_APU && adev->in_s3 && - !pm_resume_via_firmware()) { - adev->suspend_complete = false; + !pm_resume_via_firmware()) return true; - } else { - adev->suspend_complete = true; + else return false; - } } static int soc15_asic_reset(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/soc21.c b/drivers/gpu/drm/amd/amdgpu/soc21.c index 1c07ebdc0d1f..93fbb3354720 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc21.c +++ b/drivers/gpu/drm/amd/amdgpu/soc21.c @@ -897,9 +897,10 @@ static bool soc21_need_reset_on_resume(struct amdgpu_device *adev) /* Will reset for the following suspend abort cases. * 1) Only reset dGPU side. * 2) S3 suspend got aborted and TOS is active. + * As for dGPU suspend abort cases the SOL value + * will be kept as zero at this resume point. */ - if (!(adev->flags & AMD_IS_APU) && adev->in_s3 && - !adev->suspend_complete) { + if (!(adev->flags & AMD_IS_APU) && adev->in_s3) { sol_reg1 = RREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_81); msleep(100); sol_reg2 = RREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_81); -- 2.51.0 From ea9d8863daa93f2bfd39ce820254a788b1fe0c1f Mon Sep 17 00:00:00 2001 From: Le Ma Date: Fri, 25 Oct 2024 17:43:57 +0800 Subject: [PATCH 12/16] drm/amdgpu: add generic func to check if ta fw is applicable Separated xgmi ta is required for specific APU, and driver needs parse the ta binary properly with aux xgmi ta packed. v2: make the check function more generic (Lijo) Signed-off-by: Le Ma Reviewed-by: Lijo Lazar Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 34 +++++++++++++++++++++++ drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h | 1 + 2 files changed, 35 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index abd5e980c9c7..ae24df65a3df 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -3563,6 +3563,36 @@ out: return err; } +static bool is_ta_fw_applicable(struct psp_context *psp, + const struct psp_fw_bin_desc *desc) +{ + struct amdgpu_device *adev = psp->adev; + uint32_t fw_version; + + switch (desc->fw_type) { + case TA_FW_TYPE_PSP_XGMI: + case TA_FW_TYPE_PSP_XGMI_AUX: + /* for now, AUX TA only exists on 13.0.6 ta bin, + * from v20.00.0x.14 + */ + if (amdgpu_ip_version(adev, MP0_HWIP, 0) == + IP_VERSION(13, 0, 6)) { + fw_version = le32_to_cpu(desc->fw_version); + + if (adev->flags & AMD_IS_APU && + (fw_version & 0xff) >= 0x14) + return desc->fw_type == TA_FW_TYPE_PSP_XGMI_AUX; + else + return desc->fw_type == TA_FW_TYPE_PSP_XGMI; + } + break; + default: + break; + } + + return true; +} + static int parse_ta_bin_descriptor(struct psp_context *psp, const struct psp_fw_bin_desc *desc, const struct ta_firmware_header_v2_0 *ta_hdr) @@ -3572,6 +3602,9 @@ static int parse_ta_bin_descriptor(struct psp_context *psp, if (!psp || !desc || !ta_hdr) return -EINVAL; + if (!is_ta_fw_applicable(psp, desc)) + return 0; + ucode_start_addr = (uint8_t *)ta_hdr + le32_to_cpu(desc->offset_bytes) + le32_to_cpu(ta_hdr->header.ucode_array_offset_bytes); @@ -3584,6 +3617,7 @@ static int parse_ta_bin_descriptor(struct psp_context *psp, psp->asd_context.bin_desc.start_addr = ucode_start_addr; break; case TA_FW_TYPE_PSP_XGMI: + case TA_FW_TYPE_PSP_XGMI_AUX: psp->xgmi_context.context.bin_desc.fw_version = le32_to_cpu(desc->fw_version); psp->xgmi_context.context.bin_desc.size_bytes = le32_to_cpu(desc->size_bytes); psp->xgmi_context.context.bin_desc.start_addr = ucode_start_addr; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h index 4e23419b92d4..4150ec0aa10d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h @@ -163,6 +163,7 @@ enum ta_fw_type { TA_FW_TYPE_PSP_DTM, TA_FW_TYPE_PSP_RAP, TA_FW_TYPE_PSP_SECUREDISPLAY, + TA_FW_TYPE_PSP_XGMI_AUX, TA_FW_TYPE_MAX_INDEX, }; -- 2.51.0 From 7daa0f6b2859201a851f4553bea755cec14acb41 Mon Sep 17 00:00:00 2001 From: Yang Wang Date: Fri, 25 Oct 2024 13:56:03 +0800 Subject: [PATCH 13/16] drm/amdgpu: optimize ACA log print - skip to print CE ACA log. - optimize ACA log print for MCA. Signed-off-by: Yang Wang Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c | 21 ++++++++++++++++++++- 1 file changed, 20 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c index 18ee60378727..3ca03b5e0f91 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c @@ -348,6 +348,24 @@ static bool amdgpu_mca_bank_should_update(struct amdgpu_device *adev, enum amdgp return ret; } +static bool amdgpu_mca_bank_should_dump(struct amdgpu_device *adev, enum amdgpu_mca_error_type type, + struct mca_bank_entry *entry) +{ + bool ret; + + switch (type) { + case AMDGPU_MCA_ERROR_TYPE_CE: + ret = amdgpu_mca_is_deferred_error(adev, entry->regs[MCA_REG_IDX_STATUS]); + break; + case AMDGPU_MCA_ERROR_TYPE_UE: + default: + ret = true; + break; + } + + return ret; +} + static int amdgpu_mca_smu_get_mca_set(struct amdgpu_device *adev, enum amdgpu_mca_error_type type, struct mca_bank_set *mca_set, struct ras_query_context *qctx) { @@ -373,7 +391,8 @@ static int amdgpu_mca_smu_get_mca_set(struct amdgpu_device *adev, enum amdgpu_mc amdgpu_mca_bank_set_add_entry(mca_set, &entry); - amdgpu_mca_smu_mca_bank_dump(adev, i, &entry, qctx); + if (amdgpu_mca_bank_should_dump(adev, type, &entry)) + amdgpu_mca_smu_mca_bank_dump(adev, i, &entry, qctx); } return 0; -- 2.51.0 From cb67ff6272eceb5fcb2fe3b74f0293fa0706841a Mon Sep 17 00:00:00 2001 From: Jonathan Kim Date: Tue, 22 Oct 2024 12:30:50 -0400 Subject: [PATCH 14/16] drm/amdkfd: flag per-queue reset support for gfx9 Flag KFD support for per-queue reset on GFX9 devices. Signed-off-by: Jonathan Kim Reviewed-by: Harish Kasiviswanathan Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_topology.c | 2 ++ include/uapi/linux/kfd_sysfs.h | 3 ++- 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c index 3871591c9aec..9476e30d6baa 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c @@ -1998,6 +1998,8 @@ static void kfd_topology_set_capabilities(struct kfd_topology_device *dev) if (KFD_GC_VERSION(dev->gpu) >= IP_VERSION(9, 4, 2)) dev->node_props.capability |= HSA_CAP_TRAP_DEBUG_PRECISE_MEMORY_OPERATIONS_SUPPORTED; + + dev->node_props.capability |= HSA_CAP_PER_QUEUE_RESET_SUPPORTED; } else { dev->node_props.debug_prop |= HSA_DBG_WATCH_ADDR_MASK_LO_BIT_GFX10 | HSA_DBG_WATCH_ADDR_MASK_HI_BIT; diff --git a/include/uapi/linux/kfd_sysfs.h b/include/uapi/linux/kfd_sysfs.h index 5e8d28617efa..859b8e91d4d3 100644 --- a/include/uapi/linux/kfd_sysfs.h +++ b/include/uapi/linux/kfd_sysfs.h @@ -60,7 +60,8 @@ #define HSA_CAP_FLAGS_COHERENTHOSTACCESS 0x10000000 #define HSA_CAP_TRAP_DEBUG_FIRMWARE_SUPPORTED 0x20000000 #define HSA_CAP_TRAP_DEBUG_PRECISE_ALU_OPERATIONS_SUPPORTED 0x40000000 -#define HSA_CAP_RESERVED 0x800f8000 +#define HSA_CAP_PER_QUEUE_RESET_SUPPORTED 0x80000000 +#define HSA_CAP_RESERVED 0x000f8000 /* debug_prop bits in node properties */ #define HSA_DBG_WATCH_ADDR_MASK_LO_BIT_MASK 0x0000000f -- 2.51.0 From ecfe9b237687a55d596fff0650ccc8cc455edd3f Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 23 Oct 2024 09:13:21 -0400 Subject: [PATCH 15/16] drm/amdgpu/smu13: fix profile reporting The following 3 commits landed in parallel: commit d7d2688bf4ea ("drm/amd/pm: update workload mask after the setting") commit 7a1613e47e65 ("drm/amdgpu/smu13: always apply the powersave optimization") commit 7c210ca5a2d7 ("drm/amdgpu: handle default profile on on devices without fullscreen 3D") While everything is set correctly, this caused the profile to be reported incorrectly because both the powersave and fullscreen3d bits were set in the mask and when the driver prints the profile, it looks for the first bit set. Fixes: d7d2688bf4ea ("drm/amd/pm: update workload mask after the setting") Reviewed-by: Kenneth Feng Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c index 3e2277abc754..8d25cc1f218f 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c @@ -2473,7 +2473,7 @@ static int smu_v13_0_0_set_power_profile_mode(struct smu_context *smu, DpmActivityMonitorCoeffInt_t *activity_monitor = &(activity_monitor_external.DpmActivityMonitorCoeffInt); int workload_type, ret = 0; - u32 workload_mask; + u32 workload_mask, selected_workload_mask; smu->power_profile_mode = input[size]; @@ -2540,7 +2540,7 @@ static int smu_v13_0_0_set_power_profile_mode(struct smu_context *smu, if (workload_type < 0) return -EINVAL; - workload_mask = 1 << workload_type; + selected_workload_mask = workload_mask = 1 << workload_type; /* Add optimizations for SMU13.0.0/10. Reuse the power saving profile */ if ((amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 0) && @@ -2560,7 +2560,7 @@ static int smu_v13_0_0_set_power_profile_mode(struct smu_context *smu, workload_mask, NULL); if (!ret) - smu->workload_mask = workload_mask; + smu->workload_mask = selected_workload_mask; return ret; } -- 2.51.0 From 8fe7cf58ff0e46769b86b3890d657c8996b86bc6 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 14 Oct 2024 10:51:10 -0400 Subject: [PATCH 16/16] drm/amdkfd: add an interface to query whether is KFD is active Add an interface to query whether KFD has any active queues. v2: fix build issues Acked-by: Srinivasan Shanmugam Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 9 ++++++++ drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 7 ++++++ drivers/gpu/drm/amd/amdkfd/kfd_device.c | 25 ++++++++++++++++++++++ 3 files changed, 41 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index b545940e512b..24343c312480 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -890,6 +890,15 @@ int amdgpu_amdkfd_start_sched(struct amdgpu_device *adev, uint32_t node_id) return kgd2kfd_start_sched(adev->kfd.dev, node_id); } +/* check if there are KFD queues active */ +bool amdgpu_amdkfd_compute_active(struct amdgpu_device *adev, uint32_t node_id) +{ + if (!adev->kfd.init_complete) + return false; + + return kgd2kfd_compute_active(adev->kfd.dev, node_id); +} + /* Config CGTT_SQ_CLK_CTRL */ int amdgpu_amdkfd_config_sq_perfmon(struct amdgpu_device *adev, uint32_t xcp_id, bool core_override_enable, bool reg_override_enable, bool perfmon_override_enable) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index 7e0a22072536..4b80ad860639 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -268,6 +268,7 @@ int amdgpu_amdkfd_start_sched(struct amdgpu_device *adev, uint32_t node_id); int amdgpu_amdkfd_stop_sched(struct amdgpu_device *adev, uint32_t node_id); int amdgpu_amdkfd_config_sq_perfmon(struct amdgpu_device *adev, uint32_t xcp_id, bool core_override_enable, bool reg_override_enable, bool perfmon_override_enable); +bool amdgpu_amdkfd_compute_active(struct amdgpu_device *adev, uint32_t node_id); /* Read user wptr from a specified user address space with page fault @@ -431,6 +432,7 @@ int kgd2kfd_check_and_lock_kfd(void); void kgd2kfd_unlock_kfd(void); int kgd2kfd_start_sched(struct kfd_dev *kfd, uint32_t node_id); int kgd2kfd_stop_sched(struct kfd_dev *kfd, uint32_t node_id); +bool kgd2kfd_compute_active(struct kfd_dev *kfd, uint32_t node_id); #else static inline int kgd2kfd_init(void) { @@ -511,5 +513,10 @@ static inline int kgd2kfd_stop_sched(struct kfd_dev *kfd, uint32_t node_id) { return 0; } + +static inline bool kgd2kfd_compute_active(struct kfd_dev *kfd, uint32_t node_id) +{ + return false; +} #endif #endif /* AMDGPU_AMDKFD_H_INCLUDED */ diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c index fad1c8f2bc83..348925254bff 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c @@ -1392,6 +1392,13 @@ void kfd_dec_compute_active(struct kfd_node *node) WARN_ONCE(count < 0, "Compute profile ref. count error"); } +static bool kfd_compute_active(struct kfd_node *node) +{ + if (atomic_read(&node->kfd->compute_profile)) + return true; + return false; +} + void kgd2kfd_smi_event_throttle(struct kfd_dev *kfd, uint64_t throttle_bitmask) { /* @@ -1485,6 +1492,24 @@ int kgd2kfd_stop_sched(struct kfd_dev *kfd, uint32_t node_id) return node->dqm->ops.halt(node->dqm); } +bool kgd2kfd_compute_active(struct kfd_dev *kfd, uint32_t node_id) +{ + struct kfd_node *node; + + if (!kfd->init_complete) + return false; + + if (node_id >= kfd->num_nodes) { + dev_warn(kfd->adev->dev, "Invalid node ID: %u exceeds %u\n", + node_id, kfd->num_nodes - 1); + return false; + } + + node = kfd->nodes[node_id]; + + return kfd_compute_active(node); +} + #if defined(CONFIG_DEBUG_FS) /* This function will send a package to HIQ to hang the HWS -- 2.51.0