From 11510e67d0bd956878ab4ffa03c45766788092c1 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 8 Jan 2025 15:17:12 -0500 Subject: drm/amdgpu/smu13: update powersave optimizations Only apply when compute profile is selected. This is the only supported configuration. Selecting other profiles can lead to performane degradations. Reviewed-by: Kenneth Feng Signed-off-by: Alex Deucher (cherry picked from commit d477e39532d725b1cdb3c8005c689c74ffbf3b94) Cc: stable@vger.kernel.org # 6.12.x --- drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 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 2a09b27788e8..0551a3311217 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 @@ -2643,11 +2643,12 @@ static int smu_v13_0_0_set_power_profile_mode(struct smu_context *smu, &backend_workload_mask); /* 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) && - ((smu->adev->pm.fw_version == 0x004e6601) || - (smu->adev->pm.fw_version >= 0x004e7300))) || - (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 10) && - smu->adev->pm.fw_version >= 0x00504500)) { + if ((workload_mask & (1 << PP_SMC_POWER_PROFILE_COMPUTE)) && + ((amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 0) && + ((smu->adev->pm.fw_version == 0x004e6601) || + (smu->adev->pm.fw_version >= 0x004e7300))) || + (amdgpu_ip_version(smu->adev, MP1_HWIP, 0) == IP_VERSION(13, 0, 10) && + smu->adev->pm.fw_version >= 0x00504500))) { workload_type = smu_cmn_to_asic_specific_index(smu, CMN2ASIC_MAPPING_WORKLOAD, PP_SMC_POWER_PROFILE_POWERSAVING); -- cgit v1.2.3 From b5cd418f016fb801be413fd52fe4711d2d13018c Mon Sep 17 00:00:00 2001 From: Wayne Lin Date: Tue, 10 Dec 2024 11:17:55 +0800 Subject: drm/amd/display: Validate mdoe under MST LCT=1 case as well [Why & How] Currently in dm_dp_mst_is_port_support_mode(), when valdidating mode under dsc decoding at the last DP link config, we only validate the case when there is an UFP. However, if the MSTB LCT=1, there is no UFP. Under this case, use root_link_bw_in_kbps as the available bw to compare. Link: https://gitlab.freedesktop.org/drm/amd/-/issues/3720 Fixes: fa57924c76d9 ("drm/amd/display: Refactor function dm_dp_mst_is_port_support_mode()") Cc: Mario Limonciello Cc: Alex Deucher Reviewed-by: Jerry Zuo Signed-off-by: Wayne Lin Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit a04d9534a8a75b2806c5321c387be450c364b55e) Cc: stable@vger.kernel.org --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c index 6e4359490613..aadaa61ac5ac 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c @@ -1835,11 +1835,15 @@ enum dc_status dm_dp_mst_is_port_support_mode( if (immediate_upstream_port) { virtual_channel_bw_in_kbps = kbps_from_pbn(immediate_upstream_port->full_pbn); virtual_channel_bw_in_kbps = min(root_link_bw_in_kbps, virtual_channel_bw_in_kbps); - if (bw_range.min_kbps > virtual_channel_bw_in_kbps) { - DRM_DEBUG_DRIVER("MST_DSC dsc decode at last link." - "Max dsc compression can't fit into MST available bw\n"); - return DC_FAIL_BANDWIDTH_VALIDATE; - } + } else { + /* For topology LCT 1 case - only one mstb*/ + virtual_channel_bw_in_kbps = root_link_bw_in_kbps; + } + + if (bw_range.min_kbps > virtual_channel_bw_in_kbps) { + DRM_DEBUG_DRIVER("MST_DSC dsc decode at last link." + "Max dsc compression can't fit into MST available bw\n"); + return DC_FAIL_BANDWIDTH_VALIDATE; } } -- cgit v1.2.3 From adb4998f4928a17d91be054218a902ba9f8c1f93 Mon Sep 17 00:00:00 2001 From: Wayne Lin Date: Mon, 9 Dec 2024 15:25:35 +0800 Subject: drm/amd/display: Reduce accessing remote DPCD overhead [Why] Observed frame rate get dropped by tool like glxgear. Even though the output to monitor is 60Hz, the rendered frame rate drops to 30Hz lower. It's due to code path in some cases will trigger dm_dp_mst_is_port_support_mode() to read out remote Link status to assess the available bandwidth for dsc maniplation. Overhead of keep reading remote DPCD is considerable. [How] Store the remote link BW in mst_local_bw and use end-to-end full_pbn as an indicator to decide whether update the remote link bw or not. Whenever we need the info to assess the BW, visit the stored one first. Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3720 Fixes: fa57924c76d9 ("drm/amd/display: Refactor function dm_dp_mst_is_port_support_mode()") Cc: Mario Limonciello Cc: Alex Deucher Reviewed-by: Jerry Zuo Signed-off-by: Wayne Lin Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit 4a9a918545455a5979c6232fcf61ed3d8f0db3ae) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 2 ++ .../amd/display/amdgpu_dm/amdgpu_dm_mst_types.c | 34 ++++++++++++++++------ 2 files changed, 27 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h index 6464a8378387..2227cd8e4a89 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -697,6 +697,8 @@ struct amdgpu_dm_connector { struct drm_dp_mst_port *mst_output_port; struct amdgpu_dm_connector *mst_root; struct drm_dp_aux *dsc_aux; + uint32_t mst_local_bw; + uint16_t vc_full_pbn; struct mutex handle_mst_msg_ready; /* TODO see if we can merge with ddc_bus or make a dm_connector */ diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c index aadaa61ac5ac..1080075ccb17 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c @@ -155,6 +155,17 @@ amdgpu_dm_mst_connector_late_register(struct drm_connector *connector) return 0; } + +static inline void +amdgpu_dm_mst_reset_mst_connector_setting(struct amdgpu_dm_connector *aconnector) +{ + aconnector->drm_edid = NULL; + aconnector->dsc_aux = NULL; + aconnector->mst_output_port->passthrough_aux = NULL; + aconnector->mst_local_bw = 0; + aconnector->vc_full_pbn = 0; +} + static void amdgpu_dm_mst_connector_early_unregister(struct drm_connector *connector) { @@ -182,9 +193,7 @@ amdgpu_dm_mst_connector_early_unregister(struct drm_connector *connector) dc_sink_release(dc_sink); aconnector->dc_sink = NULL; - aconnector->drm_edid = NULL; - aconnector->dsc_aux = NULL; - port->passthrough_aux = NULL; + amdgpu_dm_mst_reset_mst_connector_setting(aconnector); } aconnector->mst_status = MST_STATUS_DEFAULT; @@ -504,9 +513,7 @@ dm_dp_mst_detect(struct drm_connector *connector, dc_sink_release(aconnector->dc_sink); aconnector->dc_sink = NULL; - aconnector->drm_edid = NULL; - aconnector->dsc_aux = NULL; - port->passthrough_aux = NULL; + amdgpu_dm_mst_reset_mst_connector_setting(aconnector); amdgpu_dm_set_mst_status(&aconnector->mst_status, MST_REMOTE_EDID | MST_ALLOCATE_NEW_PAYLOAD | MST_CLEAR_ALLOCATED_PAYLOAD, @@ -1819,9 +1826,18 @@ enum dc_status dm_dp_mst_is_port_support_mode( struct drm_dp_mst_port *immediate_upstream_port = NULL; uint32_t end_link_bw = 0; - /*Get last DP link BW capability*/ - if (dp_get_link_current_set_bw(&aconnector->mst_output_port->aux, &end_link_bw)) { - if (stream_kbps > end_link_bw) { + /*Get last DP link BW capability. Mode shall be supported by Legacy peer*/ + if (aconnector->mst_output_port->pdt != DP_PEER_DEVICE_DP_LEGACY_CONV && + aconnector->mst_output_port->pdt != DP_PEER_DEVICE_NONE) { + if (aconnector->vc_full_pbn != aconnector->mst_output_port->full_pbn) { + dp_get_link_current_set_bw(&aconnector->mst_output_port->aux, &end_link_bw); + aconnector->vc_full_pbn = aconnector->mst_output_port->full_pbn; + aconnector->mst_local_bw = end_link_bw; + } else { + end_link_bw = aconnector->mst_local_bw; + } + + if (end_link_bw > 0 && stream_kbps > end_link_bw) { DRM_DEBUG_DRIVER("MST_DSC dsc decode at last link." "Mode required bw can't fit into last link\n"); return DC_FAIL_BANDWIDTH_VALIDATE; -- cgit v1.2.3 From 3412860cc4c0c484f53f91b371483e6e4440c3e5 Mon Sep 17 00:00:00 2001 From: Nicholas Susanto Date: Thu, 19 Dec 2024 14:15:37 -0500 Subject: Revert "drm/amd/display: Enable urgent latency adjustments for DCN35" Revert commit 284f141f5ce5 ("drm/amd/display: Enable urgent latency adjustments for DCN35") [Why & How] Urgent latency increase caused 2.8K OLED monitor caused it to block this panel support P0. Reverting this change does not reintroduce the netflix corruption issue which it fixed. Fixes: 284f141f5ce5 ("drm/amd/display: Enable urgent latency adjustments for DCN35") Reviewed-by: Charlene Liu Signed-off-by: Nicholas Susanto Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit c7ccfc0d4241a834c25a9a9e1e78b388b4445d23) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c index beed7adbbd43..47d785204f29 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c @@ -195,9 +195,9 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc = { .dcn_downspread_percent = 0.5, .gpuvm_min_page_size_bytes = 4096, .hostvm_min_page_size_bytes = 4096, - .do_urgent_latency_adjustment = 1, + .do_urgent_latency_adjustment = 0, .urgent_latency_adjustment_fabric_clock_component_us = 0, - .urgent_latency_adjustment_fabric_clock_reference_mhz = 3000, + .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, }; void dcn35_build_wm_range_table_fpu(struct clk_mgr *clk_mgr) -- cgit v1.2.3 From ff2e4d874726c549130308b6b46aa0f8a34e04cb Mon Sep 17 00:00:00 2001 From: Leo Li Date: Mon, 9 Dec 2024 12:58:33 -0500 Subject: drm/amd/display: Do not wait for PSR disable on vbl enable [Why] Outside of a modeset/link configuration change, we should not have to wait for the panel to exit PSR. Depending on the panel and it's state, it may take multiple frames for it to exit PSR. Therefore, waiting in all scenarios may cause perceived stuttering, especially in combination with faster vblank shutdown. [How] PSR1 disable is hooked up to the vblank enable event, and vice versa. In case of vblank enable, do not wait for panel to exit PSR, but still wait in all other cases. We also avoid a call to unnecessarily change power_opts on disable - this ends up sending another command to dmcub fw. When testing against IGT, some crc tests like kms_plane_alpha_blend and amd_hotplug were failing due to CRC timeouts. This was found to be caused by the early return before HW has fully exited PSR1. Fix this by first making sure we grab a vblank reference, then waiting for panel to exit PSR1, before programming hw for CRC generation. Fixes: 58a261bfc967 ("drm/amd/display: use a more lax vblank enable policy for older ASICs") Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3743 Reviewed-by: Tom Chung Signed-off-by: Leo Li Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit aa6713fa2046f4c09bf3013dd1420ae15603ca6f) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 4 +-- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c | 25 ++++++++++------ .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c | 2 +- .../drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c | 2 +- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c | 35 ++++++++++++++++++++-- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h | 3 +- 6 files changed, 54 insertions(+), 17 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index cd16dae534dc..56b47e02db0b 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -9100,7 +9100,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, acrtc_state->stream->link->psr_settings.psr_dirty_rects_change_timestamp_ns = timestamp_ns; if (acrtc_state->stream->link->psr_settings.psr_allow_active) - amdgpu_dm_psr_disable(acrtc_state->stream); + amdgpu_dm_psr_disable(acrtc_state->stream, true); mutex_unlock(&dm->dc_lock); } } @@ -9270,7 +9270,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, if (acrtc_state->stream->link->replay_settings.replay_allow_active) amdgpu_dm_replay_disable(acrtc_state->stream); if (acrtc_state->stream->link->psr_settings.psr_allow_active) - amdgpu_dm_psr_disable(acrtc_state->stream); + amdgpu_dm_psr_disable(acrtc_state->stream, true); } mutex_unlock(&dm->dc_lock); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c index f936a35fa9eb..0f6ba7b1575d 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c @@ -30,6 +30,7 @@ #include "amdgpu_dm.h" #include "dc.h" #include "amdgpu_securedisplay.h" +#include "amdgpu_dm_psr.h" static const char *const pipe_crc_sources[] = { "none", @@ -224,6 +225,10 @@ int amdgpu_dm_crtc_configure_crc_source(struct drm_crtc *crtc, mutex_lock(&adev->dm.dc_lock); + /* For PSR1, check that the panel has exited PSR */ + if (stream_state->link->psr_settings.psr_version < DC_PSR_VERSION_SU_1) + amdgpu_dm_psr_wait_disable(stream_state); + /* Enable or disable CRTC CRC generation */ if (dm_is_crc_source_crtc(source) || source == AMDGPU_DM_PIPE_CRC_SOURCE_NONE) { if (!dc_stream_configure_crc(stream_state->ctx->dc, @@ -357,6 +362,17 @@ int amdgpu_dm_crtc_set_crc_source(struct drm_crtc *crtc, const char *src_name) } + /* + * Reading the CRC requires the vblank interrupt handler to be + * enabled. Keep a reference until CRC capture stops. + */ + enabled = amdgpu_dm_is_valid_crc_source(cur_crc_src); + if (!enabled && enable) { + ret = drm_crtc_vblank_get(crtc); + if (ret) + goto cleanup; + } + #if defined(CONFIG_DRM_AMD_SECURE_DISPLAY) /* Reset secure_display when we change crc source from debugfs */ amdgpu_dm_set_crc_window_default(crtc, crtc_state->stream); @@ -367,16 +383,7 @@ int amdgpu_dm_crtc_set_crc_source(struct drm_crtc *crtc, const char *src_name) goto cleanup; } - /* - * Reading the CRC requires the vblank interrupt handler to be - * enabled. Keep a reference until CRC capture stops. - */ - enabled = amdgpu_dm_is_valid_crc_source(cur_crc_src); if (!enabled && enable) { - ret = drm_crtc_vblank_get(crtc); - if (ret) - goto cleanup; - if (dm_is_crc_source_dprx(source)) { if (drm_dp_start_crc(aux, crtc)) { DRM_DEBUG_DRIVER("dp start crc failed\n"); 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 64a041c2af05..3011ea709b91 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 @@ -142,7 +142,7 @@ static void amdgpu_dm_crtc_set_panel_sr_feature( amdgpu_dm_replay_enable(vblank_work->stream, true); } else if (vblank_enabled) { if (link->psr_settings.psr_version < DC_PSR_VERSION_SU_1 && is_sr_active) - amdgpu_dm_psr_disable(vblank_work->stream); + amdgpu_dm_psr_disable(vblank_work->stream, false); } else if (link->psr_settings.psr_feature_enabled && allow_sr_entry && !is_sr_active && !is_crc_window_active) { diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c index 6a97bb2d9160..8b7c1a428f64 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c @@ -3606,7 +3606,7 @@ static int crc_win_update_set(void *data, u64 val) /* PSR may write to OTG CRC window control register, * so close it before starting secure_display. */ - amdgpu_dm_psr_disable(acrtc->dm_irq_params.stream); + amdgpu_dm_psr_disable(acrtc->dm_irq_params.stream, true); spin_lock_irq(&adev_to_drm(adev)->event_lock); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c index f40240aafe98..45858bf1523d 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c @@ -201,14 +201,13 @@ void amdgpu_dm_psr_enable(struct dc_stream_state *stream) * * Return: true if success */ -bool amdgpu_dm_psr_disable(struct dc_stream_state *stream) +bool amdgpu_dm_psr_disable(struct dc_stream_state *stream, bool wait) { - unsigned int power_opt = 0; bool psr_enable = false; DRM_DEBUG_DRIVER("Disabling psr...\n"); - return dc_link_set_psr_allow_active(stream->link, &psr_enable, true, false, &power_opt); + return dc_link_set_psr_allow_active(stream->link, &psr_enable, wait, false, NULL); } /* @@ -251,3 +250,33 @@ bool amdgpu_dm_psr_is_active_allowed(struct amdgpu_display_manager *dm) return allow_active; } + +/** + * amdgpu_dm_psr_wait_disable() - Wait for eDP panel to exit PSR + * @stream: stream state attached to the eDP link + * + * Waits for a max of 500ms for the eDP panel to exit PSR. + * + * Return: true if panel exited PSR, false otherwise. + */ +bool amdgpu_dm_psr_wait_disable(struct dc_stream_state *stream) +{ + enum dc_psr_state psr_state = PSR_STATE0; + struct dc_link *link = stream->link; + int retry_count; + + if (link == NULL) + return false; + + for (retry_count = 0; retry_count <= 1000; retry_count++) { + dc_link_get_psr_state(link, &psr_state); + if (psr_state == PSR_STATE0) + break; + udelay(500); + } + + if (retry_count == 1000) + return false; + + return true; +} diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h index cd2d45c2b5ef..e2366321a3c1 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h @@ -34,8 +34,9 @@ void amdgpu_dm_set_psr_caps(struct dc_link *link); void amdgpu_dm_psr_enable(struct dc_stream_state *stream); bool amdgpu_dm_link_setup_psr(struct dc_stream_state *stream); -bool amdgpu_dm_psr_disable(struct dc_stream_state *stream); +bool amdgpu_dm_psr_disable(struct dc_stream_state *stream, bool wait); bool amdgpu_dm_psr_disable_all(struct amdgpu_display_manager *dm); bool amdgpu_dm_psr_is_active_allowed(struct amdgpu_display_manager *dm); +bool amdgpu_dm_psr_wait_disable(struct dc_stream_state *stream); #endif /* AMDGPU_DM_AMDGPU_DM_PSR_H_ */ -- cgit v1.2.3 From 35ca53b7b0f0ffd16c6675fd76abac9409cf83e0 Mon Sep 17 00:00:00 2001 From: Leo Li Date: Wed, 11 Dec 2024 12:06:24 -0500 Subject: drm/amd/display: Do not elevate mem_type change to full update [Why] There should not be any need to revalidate bandwidth on memory placement change, since the fb is expected to be pinned to DCN-accessable memory before scanout. For APU it's DRAM, and DGPU, it's VRAM. However, async flips + memory type change needs to be rejected. [How] Do not set lock_and_validation_needed on mem_type change. Instead, reject an async_flip request if the crtc's buffer(s) changed mem_type. This may fix stuttering/corruption experienced with PSR SU and PSR1 panels, if the compositor allocates fbs in both VRAM carveout and GTT and flips between them. Fixes: a7c0cad0dc06 ("drm/amd/display: ensure async flips are only accepted for fast updates") Reviewed-by: Tom Chung Signed-off-by: Leo Li Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit 4caacd1671b7a013ad04cd8b6398f002540bdd4d) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 29 ++++++++++++++++++----- 1 file changed, 23 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index 56b47e02db0b..dcc5d8ded662 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -11379,6 +11379,25 @@ static int dm_crtc_get_cursor_mode(struct amdgpu_device *adev, return 0; } +static bool amdgpu_dm_crtc_mem_type_changed(struct drm_device *dev, + struct drm_atomic_state *state, + struct drm_crtc_state *crtc_state) +{ + struct drm_plane *plane; + struct drm_plane_state *new_plane_state, *old_plane_state; + + drm_for_each_plane_mask(plane, dev, crtc_state->plane_mask) { + new_plane_state = drm_atomic_get_plane_state(state, plane); + old_plane_state = drm_atomic_get_plane_state(state, plane); + + if (old_plane_state->fb && new_plane_state->fb && + get_mem_type(old_plane_state->fb) != get_mem_type(new_plane_state->fb)) + return true; + } + + return false; +} + /** * amdgpu_dm_atomic_check() - Atomic check implementation for AMDgpu DM. * @@ -11576,10 +11595,6 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev, /* Remove exiting planes if they are modified */ for_each_oldnew_plane_in_descending_zpos(state, plane, old_plane_state, new_plane_state) { - if (old_plane_state->fb && new_plane_state->fb && - get_mem_type(old_plane_state->fb) != - get_mem_type(new_plane_state->fb)) - lock_and_validation_needed = true; ret = dm_update_plane_state(dc, state, plane, old_plane_state, @@ -11874,9 +11889,11 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev, /* * Only allow async flips for fast updates that don't change - * the FB pitch, the DCC state, rotation, etc. + * the FB pitch, the DCC state, rotation, mem_type, etc. */ - if (new_crtc_state->async_flip && lock_and_validation_needed) { + if (new_crtc_state->async_flip && + (lock_and_validation_needed || + amdgpu_dm_crtc_mem_type_changed(dev, state, new_crtc_state))) { drm_dbg_atomic(crtc->dev, "[CRTC:%d:%s] async flips are only supported for fast updates\n", crtc->base.id, crtc->name); -- cgit v1.2.3 From 470d4f05c77153b97f53554a3c44164361a4d939 Mon Sep 17 00:00:00 2001 From: Yiling Chen Date: Fri, 20 Dec 2024 17:12:54 +0800 Subject: drm/amd/display: Remove unnecessary eDP power down [why] When first time of link training is fail, eDP would be powered down and would not be powered up for next retry link training. It causes that all of retry link linking would be fail. [how] We has extracted both power up and down sequence from enable/disable link output function before DCN32. We remov eDP power down in dcn32_disable_link_output(). Reviewed-by: Charlene Liu Signed-off-by: Yiling Chen Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit f5860c88cdfe7300d08c1aef881bba0cac369e34) --- drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c index fa11f075d1f9..d9136d252ee1 100644 --- a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c @@ -1398,12 +1398,12 @@ void dcn32_disable_link_output(struct dc_link *link, link_hwss->disable_link_output(link, link_res, signal); link->phy_state.symclk_state = SYMCLK_OFF_TX_OFF; - - if (signal == SIGNAL_TYPE_EDP && - link->dc->hwss.edp_power_control && - !link->skip_implict_edp_power_control) - link->dc->hwss.edp_power_control(link, false); - else if (dmcu != NULL && dmcu->funcs->unlock_phy) + /* + * Add the logic to extract BOTH power up and power down sequences + * from enable/disable link output and only call edp panel control + * in enable_link_dp and disable_link_dp once. + */ + if (dmcu != NULL && dmcu->funcs->unlock_phy) dmcu->funcs->unlock_phy(dmcu); dc->link_srv->dp_trace_source_sequence(link, DPCD_SOURCE_SEQ_AFTER_DISABLE_LINK_PHY); -- cgit v1.2.3 From b5c764d6ed556c4e81fbe3fd976da77ec450c08e Mon Sep 17 00:00:00 2001 From: Tom Chung Date: Tue, 1 Oct 2024 17:13:07 +0800 Subject: drm/amd/display: Use HW lock mgr for PSR1 [Why] Without the dmub hw lock, it may cause the lock timeout issue while do modeset on PSR1 eDP panel. [How] Allow dmub hw lock for PSR1. Reviewed-by: Sun peng Li Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit a2b5a9956269f4c1a09537177f18ab0229fe79f7) --- drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c index bf636b28e3e1..5bb8b78bf250 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c @@ -63,7 +63,8 @@ void dmub_hw_lock_mgr_inbox0_cmd(struct dc_dmub_srv *dmub_srv, bool should_use_dmub_lock(struct dc_link *link) { - if (link->psr_settings.psr_version == DC_PSR_VERSION_SU_1) + if (link->psr_settings.psr_version == DC_PSR_VERSION_SU_1 || + link->psr_settings.psr_version == DC_PSR_VERSION_1) return true; if (link->replay_settings.replay_feature_enabled) -- cgit v1.2.3 From 36b23e3baf9129d5b6c3a3a85b6b7ffb75ae287c Mon Sep 17 00:00:00 2001 From: Alex Hung Date: Tue, 17 Dec 2024 11:51:58 -0700 Subject: drm/amd/display: Initialize denominator defaults to 1 [WHAT & HOW] Variables, used as denominators and maybe not assigned to other values, should be initialized to non-zero to avoid DIVIDE_BY_ZERO, as reported by Coverity. Reviewed-by: Austin Zheng Reviewed-by: Rodrigo Siqueira Signed-off-by: Alex Hung Signed-off-by: Tom Chung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit e2c4c6c10542ccfe4a0830bb6c9fd5b177b7bbb7) --- .../amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c index 601320b1be81..ecfa3c898e09 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c @@ -579,8 +579,8 @@ static void CalculateBytePerPixelAndBlockSizes( { *BytePerPixelDETY = 0; *BytePerPixelDETC = 0; - *BytePerPixelY = 0; - *BytePerPixelC = 0; + *BytePerPixelY = 1; + *BytePerPixelC = 1; if (SourcePixelFormat == dml2_444_64) { *BytePerPixelDETY = 8; -- cgit v1.2.3 From b0a3e840ad287c33a86b5515d606451b7df86ad4 Mon Sep 17 00:00:00 2001 From: Tom Chung Date: Thu, 5 Dec 2024 23:08:28 +0800 Subject: drm/amd/display: Fix PSR-SU not support but still call the amdgpu_dm_psr_enable [Why] The enum DC_PSR_VERSION_SU_1 of psr_version is 1 and DC_PSR_VERSION_UNSUPPORTED is 0xFFFFFFFF. The original code may has chance trigger the amdgpu_dm_psr_enable() while psr version is set to DC_PSR_VERSION_UNSUPPORTED. [How] Modify the condition to psr->psr_version == DC_PSR_VERSION_SU_1 Reviewed-by: Sun peng Li Signed-off-by: Tom Chung Signed-off-by: Roman Li Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit f765e7ce0417f8dc38479b4b495047c397c16902) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index dcc5d8ded662..a5fd19ee89dc 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -8929,7 +8929,7 @@ static void amdgpu_dm_enable_self_refresh(struct amdgpu_crtc *acrtc_attach, (current_ts - psr->psr_dirty_rects_change_timestamp_ns) > 500000000) { if (pr->replay_feature_enabled && !pr->replay_allow_active) amdgpu_dm_replay_enable(acrtc_state->stream, true); - if (psr->psr_version >= DC_PSR_VERSION_SU_1 && + if (psr->psr_version == DC_PSR_VERSION_SU_1 && !psr->psr_allow_active && !aconn->disallow_edp_enter_psr) amdgpu_dm_psr_enable(acrtc_state->stream); } -- cgit v1.2.3 From 67edb81d6e9af43a0d58edf74630f82cfda4155d Mon Sep 17 00:00:00 2001 From: Tom Chung Date: Thu, 5 Dec 2024 23:20:45 +0800 Subject: drm/amd/display: Disable replay and psr while VRR is enabled [Why] Replay and PSR will cause some video corruption while VRR is enabled. [How] 1. Disable the Replay and PSR while VRR is enabled. 2. Change the amdgpu_dm_crtc_vrr_active() parameter to const. Because the function will only read data from dm_crtc_state. Reviewed-by: Sun peng Li Signed-off-by: Tom Chung Signed-off-by: Roman Li Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher (cherry picked from commit d7879340e987b3056b8ae39db255b6c19c170a0d) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 6 ++++-- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c | 2 +- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.h | 2 +- 3 files changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index a5fd19ee89dc..5f216d626cbb 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -8896,6 +8896,7 @@ static void amdgpu_dm_enable_self_refresh(struct amdgpu_crtc *acrtc_attach, struct replay_settings *pr = &acrtc_state->stream->link->replay_settings; struct amdgpu_dm_connector *aconn = (struct amdgpu_dm_connector *)acrtc_state->stream->dm_stream_context; + bool vrr_active = amdgpu_dm_crtc_vrr_active(acrtc_state); if (acrtc_state->update_type > UPDATE_TYPE_FAST) { if (pr->config.replay_supported && !pr->replay_feature_enabled) @@ -8922,7 +8923,8 @@ static void amdgpu_dm_enable_self_refresh(struct amdgpu_crtc *acrtc_attach, * adequate number of fast atomic commits to notify KMD * of update events. See `vblank_control_worker()`. */ - if (acrtc_attach->dm_irq_params.allow_sr_entry && + if (!vrr_active && + acrtc_attach->dm_irq_params.allow_sr_entry && #ifdef CONFIG_DRM_AMD_SECURE_DISPLAY !amdgpu_dm_crc_window_is_activated(acrtc_state->base.crtc) && #endif @@ -9266,7 +9268,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, bundle->stream_update.abm_level = &acrtc_state->abm_level; mutex_lock(&dm->dc_lock); - if (acrtc_state->update_type > UPDATE_TYPE_FAST) { + if ((acrtc_state->update_type > UPDATE_TYPE_FAST) || vrr_active) { if (acrtc_state->stream->link->replay_settings.replay_allow_active) amdgpu_dm_replay_disable(acrtc_state->stream); if (acrtc_state->stream->link->psr_settings.psr_allow_active) 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 3011ea709b91..36a830a7440f 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 @@ -93,7 +93,7 @@ int amdgpu_dm_crtc_set_vupdate_irq(struct drm_crtc *crtc, bool enable) return rc; } -bool amdgpu_dm_crtc_vrr_active(struct dm_crtc_state *dm_state) +bool amdgpu_dm_crtc_vrr_active(const struct dm_crtc_state *dm_state) { return dm_state->freesync_config.state == VRR_STATE_ACTIVE_VARIABLE || dm_state->freesync_config.state == VRR_STATE_ACTIVE_FIXED; diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.h index 17e948753f59..c1212947a77b 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.h @@ -37,7 +37,7 @@ int amdgpu_dm_crtc_set_vupdate_irq(struct drm_crtc *crtc, bool enable); bool amdgpu_dm_crtc_vrr_active_irq(struct amdgpu_crtc *acrtc); -bool amdgpu_dm_crtc_vrr_active(struct dm_crtc_state *dm_state); +bool amdgpu_dm_crtc_vrr_active(const struct dm_crtc_state *dm_state); int amdgpu_dm_crtc_enable_vblank(struct drm_crtc *crtc); -- cgit v1.2.3 From 1e8c193f8ca7ab7dff4f4747b45a55dca23c00f4 Mon Sep 17 00:00:00 2001 From: Srinivasan Shanmugam Date: Thu, 9 Jan 2025 21:33:51 +0530 Subject: drm/amdgpu: Fix Circular Locking Dependency in AMDGPU GFX Isolation MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This commit addresses a circular locking dependency issue within the GFX isolation mechanism. The problem was identified by a warning indicating a potential deadlock due to inconsistent lock acquisition order. - The `amdgpu_gfx_enforce_isolation_ring_begin_use` and `amdgpu_gfx_enforce_isolation_ring_end_use` functions previously acquired `enforce_isolation_mutex` and called `amdgpu_gfx_kfd_sch_ctrl`, leading to potential deadlocks. ie., If `amdgpu_gfx_kfd_sch_ctrl` is called while `enforce_isolation_mutex` is held, and `amdgpu_gfx_enforce_isolation_handler` is called while `kfd_sch_mutex` is held, it can create a circular dependency. By ensuring consistent lock usage, this fix resolves the issue: [ 606.297333] ====================================================== [ 606.297343] WARNING: possible circular locking dependency detected [ 606.297353] 6.10.0-amd-mlkd-610-311224-lof #19 Tainted: G OE [ 606.297365] ------------------------------------------------------ [ 606.297375] kworker/u96:3/3825 is trying to acquire lock: [ 606.297385] ffff9aa64e431cb8 ((work_completion)(&(&adev->gfx.enforce_isolation[i].work)->work)){+.+.}-{0:0}, at: __flush_work+0x232/0x610 [ 606.297413] but task is already holding lock: [ 606.297423] ffff9aa64e432338 (&adev->gfx.kfd_sch_mutex){+.+.}-{3:3}, at: amdgpu_gfx_kfd_sch_ctrl+0x51/0x4d0 [amdgpu] [ 606.297725] which lock already depends on the new lock. [ 606.297738] the existing dependency chain (in reverse order) is: [ 606.297749] -> #2 (&adev->gfx.kfd_sch_mutex){+.+.}-{3:3}: [ 606.297765] __mutex_lock+0x85/0x930 [ 606.297776] mutex_lock_nested+0x1b/0x30 [ 606.297786] amdgpu_gfx_kfd_sch_ctrl+0x51/0x4d0 [amdgpu] [ 606.298007] amdgpu_gfx_enforce_isolation_ring_begin_use+0x2a4/0x5d0 [amdgpu] [ 606.298225] amdgpu_ring_alloc+0x48/0x70 [amdgpu] [ 606.298412] amdgpu_ib_schedule+0x176/0x8a0 [amdgpu] [ 606.298603] amdgpu_job_run+0xac/0x1e0 [amdgpu] [ 606.298866] drm_sched_run_job_work+0x24f/0x430 [gpu_sched] [ 606.298880] process_one_work+0x21e/0x680 [ 606.298890] worker_thread+0x190/0x350 [ 606.298899] kthread+0xe7/0x120 [ 606.298908] ret_from_fork+0x3c/0x60 [ 606.298919] ret_from_fork_asm+0x1a/0x30 [ 606.298929] -> #1 (&adev->enforce_isolation_mutex){+.+.}-{3:3}: [ 606.298947] __mutex_lock+0x85/0x930 [ 606.298956] mutex_lock_nested+0x1b/0x30 [ 606.298966] amdgpu_gfx_enforce_isolation_handler+0x87/0x370 [amdgpu] [ 606.299190] process_one_work+0x21e/0x680 [ 606.299199] worker_thread+0x190/0x350 [ 606.299208] kthread+0xe7/0x120 [ 606.299217] ret_from_fork+0x3c/0x60 [ 606.299227] ret_from_fork_asm+0x1a/0x30 [ 606.299236] -> #0 ((work_completion)(&(&adev->gfx.enforce_isolation[i].work)->work)){+.+.}-{0:0}: [ 606.299257] __lock_acquire+0x16f9/0x2810 [ 606.299267] lock_acquire+0xd1/0x300 [ 606.299276] __flush_work+0x250/0x610 [ 606.299286] cancel_delayed_work_sync+0x71/0x80 [ 606.299296] amdgpu_gfx_kfd_sch_ctrl+0x287/0x4d0 [amdgpu] [ 606.299509] amdgpu_gfx_enforce_isolation_ring_begin_use+0x2a4/0x5d0 [amdgpu] [ 606.299723] amdgpu_ring_alloc+0x48/0x70 [amdgpu] [ 606.299909] amdgpu_ib_schedule+0x176/0x8a0 [amdgpu] [ 606.300101] amdgpu_job_run+0xac/0x1e0 [amdgpu] [ 606.300355] drm_sched_run_job_work+0x24f/0x430 [gpu_sched] [ 606.300369] process_one_work+0x21e/0x680 [ 606.300378] worker_thread+0x190/0x350 [ 606.300387] kthread+0xe7/0x120 [ 606.300396] ret_from_fork+0x3c/0x60 [ 606.300406] ret_from_fork_asm+0x1a/0x30 [ 606.300416] other info that might help us debug this: [ 606.300428] Chain exists of: (work_completion)(&(&adev->gfx.enforce_isolation[i].work)->work) --> &adev->enforce_isolation_mutex --> &adev->gfx.kfd_sch_mutex [ 606.300458] Possible unsafe locking scenario: [ 606.300468] CPU0 CPU1 [ 606.300476] ---- ---- [ 606.300484] lock(&adev->gfx.kfd_sch_mutex); [ 606.300494] lock(&adev->enforce_isolation_mutex); [ 606.300508] lock(&adev->gfx.kfd_sch_mutex); [ 606.300521] lock((work_completion)(&(&adev->gfx.enforce_isolation[i].work)->work)); [ 606.300536] *** DEADLOCK *** [ 606.300546] 5 locks held by kworker/u96:3/3825: [ 606.300555] #0: ffff9aa5aa1f5d58 ((wq_completion)comp_1.1.0){+.+.}-{0:0}, at: process_one_work+0x3f5/0x680 [ 606.300577] #1: ffffaa53c3c97e40 ((work_completion)(&sched->work_run_job)){+.+.}-{0:0}, at: process_one_work+0x1d6/0x680 [ 606.300600] #2: ffff9aa64e463c98 (&adev->enforce_isolation_mutex){+.+.}-{3:3}, at: amdgpu_gfx_enforce_isolation_ring_begin_use+0x1c3/0x5d0 [amdgpu] [ 606.300837] #3: ffff9aa64e432338 (&adev->gfx.kfd_sch_mutex){+.+.}-{3:3}, at: amdgpu_gfx_kfd_sch_ctrl+0x51/0x4d0 [amdgpu] [ 606.301062] #4: ffffffff8c1a5660 (rcu_read_lock){....}-{1:2}, at: __flush_work+0x70/0x610 [ 606.301083] stack backtrace: [ 606.301092] CPU: 14 PID: 3825 Comm: kworker/u96:3 Tainted: G OE 6.10.0-amd-mlkd-610-311224-lof #19 [ 606.301109] Hardware name: Gigabyte Technology Co., Ltd. X570S GAMING X/X570S GAMING X, BIOS F7 03/22/2024 [ 606.301124] Workqueue: comp_1.1.0 drm_sched_run_job_work [gpu_sched] [ 606.301140] Call Trace: [ 606.301146] [ 606.301154] dump_stack_lvl+0x9b/0xf0 [ 606.301166] dump_stack+0x10/0x20 [ 606.301175] print_circular_bug+0x26c/0x340 [ 606.301187] check_noncircular+0x157/0x170 [ 606.301197] ? register_lock_class+0x48/0x490 [ 606.301213] __lock_acquire+0x16f9/0x2810 [ 606.301230] lock_acquire+0xd1/0x300 [ 606.301239] ? __flush_work+0x232/0x610 [ 606.301250] ? srso_alias_return_thunk+0x5/0xfbef5 [ 606.301261] ? mark_held_locks+0x54/0x90 [ 606.301274] ? __flush_work+0x232/0x610 [ 606.301284] __flush_work+0x250/0x610 [ 606.301293] ? __flush_work+0x232/0x610 [ 606.301305] ? __pfx_wq_barrier_func+0x10/0x10 [ 606.301318] ? mark_held_locks+0x54/0x90 [ 606.301331] ? srso_alias_return_thunk+0x5/0xfbef5 [ 606.301345] cancel_delayed_work_sync+0x71/0x80 [ 606.301356] amdgpu_gfx_kfd_sch_ctrl+0x287/0x4d0 [amdgpu] [ 606.301661] amdgpu_gfx_enforce_isolation_ring_begin_use+0x2a4/0x5d0 [amdgpu] [ 606.302050] ? srso_alias_return_thunk+0x5/0xfbef5 [ 606.302069] amdgpu_ring_alloc+0x48/0x70 [amdgpu] [ 606.302452] amdgpu_ib_schedule+0x176/0x8a0 [amdgpu] [ 606.302862] ? drm_sched_entity_error+0x82/0x190 [gpu_sched] [ 606.302890] amdgpu_job_run+0xac/0x1e0 [amdgpu] [ 606.303366] drm_sched_run_job_work+0x24f/0x430 [gpu_sched] [ 606.303388] process_one_work+0x21e/0x680 [ 606.303409] worker_thread+0x190/0x350 [ 606.303424] ? __pfx_worker_thread+0x10/0x10 [ 606.303437] kthread+0xe7/0x120 [ 606.303449] ? __pfx_kthread+0x10/0x10 [ 606.303463] ret_from_fork+0x3c/0x60 [ 606.303476] ? __pfx_kthread+0x10/0x10 [ 606.303489] ret_from_fork_asm+0x1a/0x30 [ 606.303512] v2: Refactor lock handling to resolve circular dependency (Alex) - Introduced a `sched_work` flag to defer the call to `amdgpu_gfx_kfd_sch_ctrl` until after releasing `enforce_isolation_mutex`. - This change ensures that `amdgpu_gfx_kfd_sch_ctrl` is called outside the critical section, preventing the circular dependency and deadlock. - The `sched_work` flag is set within the mutex-protected section if conditions are met, and the actual function call is made afterward. - This approach ensures consistent lock acquisition order. Fixes: afefd6f24502 ("drm/amdgpu: Implement Enforce Isolation Handler for KGD/KFD serialization") Cc: Christian König Cc: Alex Deucher Signed-off-by: Srinivasan Shanmugam Suggested-by: Alex Deucher Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher (cherry picked from commit 0b6b2dd38336d5fd49214f0e4e6495e658e3ab44) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c index 69a6b6dba0a5..1d155463d044 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c @@ -1989,6 +1989,7 @@ void amdgpu_gfx_enforce_isolation_ring_begin_use(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; u32 idx; + bool sched_work = false; if (!adev->gfx.enable_cleaner_shader) return; @@ -2007,15 +2008,19 @@ void amdgpu_gfx_enforce_isolation_ring_begin_use(struct amdgpu_ring *ring) mutex_lock(&adev->enforce_isolation_mutex); if (adev->enforce_isolation[idx]) { if (adev->kfd.init_complete) - amdgpu_gfx_kfd_sch_ctrl(adev, idx, false); + sched_work = true; } mutex_unlock(&adev->enforce_isolation_mutex); + + if (sched_work) + amdgpu_gfx_kfd_sch_ctrl(adev, idx, false); } void amdgpu_gfx_enforce_isolation_ring_end_use(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; u32 idx; + bool sched_work = false; if (!adev->gfx.enable_cleaner_shader) return; @@ -2031,9 +2036,12 @@ void amdgpu_gfx_enforce_isolation_ring_end_use(struct amdgpu_ring *ring) mutex_lock(&adev->enforce_isolation_mutex); if (adev->enforce_isolation[idx]) { if (adev->kfd.init_complete) - amdgpu_gfx_kfd_sch_ctrl(adev, idx, true); + sched_work = true; } mutex_unlock(&adev->enforce_isolation_mutex); + + if (sched_work) + amdgpu_gfx_kfd_sch_ctrl(adev, idx, true); } /* -- cgit v1.2.3 From 90505894c4ed581318836b792c57723df491cb91 Mon Sep 17 00:00:00 2001 From: Kenneth Feng Date: Thu, 9 Jan 2025 15:58:23 +0800 Subject: drm/amdgpu: disable gfxoff with the compute workload on gfx12 Disable gfxoff with the compute workload on gfx12. This is a workaround for the opencl test failure. Signed-off-by: Kenneth Feng Acked-by: Alex Deucher Signed-off-by: Alex Deucher (cherry picked from commit 2affe2bbc997b3920045c2c434e480c81a5f9707) Cc: stable@vger.kernel.org # 6.12.x --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index 3afcd1e8aa54..c4e733c2e75e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -715,8 +715,9 @@ err: void amdgpu_amdkfd_set_compute_idle(struct amdgpu_device *adev, bool idle) { enum amd_powergating_state state = idle ? AMD_PG_STATE_GATE : AMD_PG_STATE_UNGATE; - if (IP_VERSION_MAJ(amdgpu_ip_version(adev, GC_HWIP, 0)) == 11 && - ((adev->mes.kiq_version & AMDGPU_MES_VERSION_MASK) <= 64)) { + if ((IP_VERSION_MAJ(amdgpu_ip_version(adev, GC_HWIP, 0)) == 11 && + ((adev->mes.kiq_version & AMDGPU_MES_VERSION_MASK) <= 64)) || + (IP_VERSION_MAJ(amdgpu_ip_version(adev, GC_HWIP, 0)) == 12)) { pr_debug("GFXOFF is %s\n", idle ? "enabled" : "disabled"); amdgpu_gfx_off_ctrl(adev, idle); } else if ((IP_VERSION_MAJ(amdgpu_ip_version(adev, GC_HWIP, 0)) == 9) && -- cgit v1.2.3 From af04b320c71c4b59971f021615876808a36e5038 Mon Sep 17 00:00:00 2001 From: Christian König Date: Fri, 20 Dec 2024 16:21:11 +0100 Subject: drm/amdgpu: always sync the GFX pipe on ctx switch MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit That is needed to enforce isolation between contexts. Signed-off-by: Christian König Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher (cherry picked from commit def59436fb0d3ca0f211d14873d0273d69ebb405) Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index 8b512dc28df8..071f187f5e28 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -193,8 +193,8 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned int num_ibs, need_ctx_switch = ring->current_ctx != fence_ctx; if (ring->funcs->emit_pipeline_sync && job && ((tmp = amdgpu_sync_get_fence(&job->explicit_sync)) || - (amdgpu_sriov_vf(adev) && need_ctx_switch) || - amdgpu_vm_need_pipeline_sync(ring, job))) { + need_ctx_switch || amdgpu_vm_need_pipeline_sync(ring, job))) { + need_pipe_sync = true; if (tmp) -- cgit v1.2.3 From bd275e6cfc972329d39c6406a3c6d2ba2aba7db6 Mon Sep 17 00:00:00 2001 From: Gui Chengming Date: Tue, 7 Jan 2025 17:09:08 +0800 Subject: drm/amdgpu: fix fw attestation for MP0_14_0_{2/3} MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit FW attestation was disabled on MP0_14_0_{2/3}. V2: Move check into is_fw_attestation_support func. (Frank) Remove DRM_WARN log info. (Alex) Fix format. (Christian) Signed-off-by: Gui Chengming Reviewed-by: Frank.Min Reviewed-by: Christian König Signed-off-by: Alex Deucher (cherry picked from commit 62952a38d9bcf357d5ffc97615c48b12c9cd627c) Cc: stable@vger.kernel.org # 6.12.x --- drivers/gpu/drm/amd/amdgpu/amdgpu_fw_attestation.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fw_attestation.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fw_attestation.c index 2d4b67175b55..328a1b963548 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fw_attestation.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fw_attestation.c @@ -122,6 +122,10 @@ static int amdgpu_is_fw_attestation_supported(struct amdgpu_device *adev) if (adev->flags & AMD_IS_APU) return 0; + if (amdgpu_ip_version(adev, MP0_HWIP, 0) == IP_VERSION(14, 0, 2) || + amdgpu_ip_version(adev, MP0_HWIP, 0) == IP_VERSION(14, 0, 3)) + return 0; + if (adev->asic_type >= CHIP_SIENNA_CICHLID) return 1; -- cgit v1.2.3