From f214b7beb00621b983e67ce97477afc3ab4b38f4 Mon Sep 17 00:00:00 2001 From: Jay Cornwall Date: Thu, 16 Jan 2025 14:36:39 -0600 Subject: [PATCH 1/6] drm/amdkfd: Block per-queue reset when halt_if_hws_hang=1 The purpose of halt_if_hws_hang is to preserve GPU state for driver debugging when queue preemption fails. Issuing per-queue reset may kill wavefronts which caused the preemption failure. Signed-off-by: Jay Cornwall Reviewed-by: Jonathan Kim Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org # 6.12.x --- drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c index 1405e8affd48..d4593374e7a1 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -2325,9 +2325,9 @@ static int unmap_queues_cpsch(struct device_queue_manager *dqm, */ mqd_mgr = dqm->mqd_mgrs[KFD_MQD_TYPE_HIQ]; if (mqd_mgr->check_preemption_failed(mqd_mgr, dqm->packet_mgr.priv_queue->queue->mqd)) { + while (halt_if_hws_hang) + schedule(); if (reset_queues_on_hws_hang(dqm)) { - while (halt_if_hws_hang) - schedule(); dqm->is_hws_hang = true; kfd_hws_hang(dqm); retval = -ETIME; From 5cda56bd86c455341087dca29c65dc7c87f84340 Mon Sep 17 00:00:00 2001 From: Kenneth Feng Date: Mon, 20 Jan 2025 15:33:03 +0800 Subject: [PATCH 2/6] drm/amd/amdgpu: change the config of cgcg on gfx12 change the config of cgcg on gfx12 Signed-off-by: Kenneth Feng Reviewed-by: Yang Wang Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org # 6.12.x --- drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c | 11 ----------- 1 file changed, 11 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c index c1772f44b1d7..2523221a2519 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c @@ -4021,17 +4021,6 @@ static void gfx_v12_0_update_coarse_grain_clock_gating(struct amdgpu_device *ade if (def != data) WREG32_SOC15(GC, 0, regRLC_CGCG_CGLS_CTRL_3D, data); - - data = RREG32_SOC15(GC, 0, regSDMA0_RLC_CGCG_CTRL); - data &= ~SDMA0_RLC_CGCG_CTRL__CGCG_INT_ENABLE_MASK; - WREG32_SOC15(GC, 0, regSDMA0_RLC_CGCG_CTRL, data); - - /* Some ASICs only have one SDMA instance, not need to configure SDMA1 */ - if (adev->sdma.num_instances > 1) { - data = RREG32_SOC15(GC, 0, regSDMA1_RLC_CGCG_CTRL); - data &= ~SDMA1_RLC_CGCG_CTRL__CGCG_INT_ENABLE_MASK; - WREG32_SOC15(GC, 0, regSDMA1_RLC_CGCG_CTRL, data); - } } } From 819bf6662b93a5a8b0c396d2c7e7fab6264c9808 Mon Sep 17 00:00:00 2001 From: Lijo Lazar Date: Wed, 22 Jan 2025 09:12:41 +0530 Subject: [PATCH 3/6] drm/amd/pm: Mark MM activity as unsupported Aldebaran doesn't support querying MM activity percentage. Keep the field as 0xFFs to mark it as unsupported. Signed-off-by: Lijo Lazar Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c index f6b029354327..83163d7c7f00 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c @@ -1732,7 +1732,6 @@ static ssize_t aldebaran_get_gpu_metrics(struct smu_context *smu, gpu_metrics->average_gfx_activity = metrics.AverageGfxActivity; gpu_metrics->average_umc_activity = metrics.AverageUclkActivity; - gpu_metrics->average_mm_activity = 0; /* Valid power data is available only from primary die */ if (aldebaran_is_primary(smu)) { From f88192d2335b5a911fcfa09338cc00624571ec5e Mon Sep 17 00:00:00 2001 From: loanchen Date: Wed, 15 Jan 2025 17:43:29 +0800 Subject: [PATCH 4/6] drm/amd/display: Correct register address in dcn35 [Why] the offset address of mmCLK5_spll_field_8 was incorrect for dcn35 which causes SSC not to be enabled. Reviewed-by: Charlene Liu Signed-off-by: Lo-An Chen Signed-off-by: Zaeem Mohamed Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) 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 1f974ea3b0c6..1648226586e2 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 @@ -89,7 +89,7 @@ #define mmCLK1_CLK4_ALLOW_DS 0x16EA8 #define mmCLK1_CLK5_ALLOW_DS 0x16EB1 -#define mmCLK5_spll_field_8 0x1B04B +#define mmCLK5_spll_field_8 0x1B24B #define mmDENTIST_DISPCLK_CNTL 0x0124 #define regDENTIST_DISPCLK_CNTL 0x0064 #define regDENTIST_DISPCLK_CNTL_BASE_IDX 1 From 9078a5bfa21e78ae68b6d7c365d1b92f26720c55 Mon Sep 17 00:00:00 2001 From: Prike Liang Date: Tue, 14 Jan 2025 11:20:17 +0800 Subject: [PATCH 5/6] drm/amdkfd: only flush the validate MES contex The following page fault was observed duringthe KFD process release. In this particular error case, the HIP test (./MemcpyPerformance -h) does not require the queue. As a result, the process_context_addr was not assigned when the KFD process was released, ultimately leading to this page fault during the execution of the function kfd_process_dequeue_from_all_devices(). [345962.294891] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:153 vmid:0 pasid:0) [345962.295333] amdgpu 0000:03:00.0: amdgpu: in page starting at address 0x0000000000000000 from client 10 [345962.295775] amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00000B33 [345962.296097] amdgpu 0000:03:00.0: amdgpu: Faulty UTCL2 client ID: CPC (0x5) [345962.296394] amdgpu 0000:03:00.0: amdgpu: MORE_FAULTS: 0x1 [345962.296633] amdgpu 0000:03:00.0: amdgpu: WALKER_ERROR: 0x1 [345962.296876] amdgpu 0000:03:00.0: amdgpu: PERMISSION_FAULTS: 0x3 [345962.297135] amdgpu 0000:03:00.0: amdgpu: MAPPING_ERROR: 0x1 [345962.297377] amdgpu 0000:03:00.0: amdgpu: RW: 0x0 [345962.297682] amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:169 vmid:0 pasid:0) Signed-off-by: Prike Liang Reviewed-by: Jonathan Kim Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c index 9df56f8e09f9..bcddd989c7f3 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c @@ -86,9 +86,12 @@ void kfd_process_dequeue_from_device(struct kfd_process_device *pdd) if (pdd->already_dequeued) return; - + /* The MES context flush needs to filter out the case which the + * KFD process is created without setting up the MES context and + * queue for creating a compute queue. + */ dev->dqm->ops.process_termination(dev->dqm, &pdd->qpd); - if (dev->kfd->shared_resources.enable_mes && + if (dev->kfd->shared_resources.enable_mes && !!pdd->proc_ctx_gpu_addr && down_read_trylock(&dev->adev->reset_domain->sem)) { amdgpu_mes_flush_shader_debugger(dev->adev, pdd->proc_ctx_gpu_addr); From 7f2b5237e313e39008a85b33ca94ab503a8fdff9 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Mon, 27 Jan 2025 21:41:10 -0300 Subject: [PATCH 6/6] drm/amd/display: restore invalid MSA timing check for freesync This restores the original behavior that gets min/max freq from EDID and only set DP/eDP connector as freesync capable if "sink device is capable of rendering incoming video stream without MSA timing parameters", i.e., `allow_invalid_MSA_timing_params` is true. The condition was mistakenly removed by 0159f88a99c9 ("drm/amd/display: remove redundant freesync parser for DP"). CC: Mario Limonciello CC: Alex Hung Link: https://gitlab.freedesktop.org/drm/amd/-/issues/3915 Fixes: 0159f88a99c9 ("drm/amd/display: remove redundant freesync parser for DP") Reviewed-by: Harry Wentland Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 12 ++++++++---- 1 file changed, 8 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 b0e66c05d811..ac3fd81fecef 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -12326,10 +12326,14 @@ void amdgpu_dm_update_freesync_caps(struct drm_connector *connector, if (edid && (sink->sink_signal == SIGNAL_TYPE_DISPLAY_PORT || sink->sink_signal == SIGNAL_TYPE_EDP)) { - amdgpu_dm_connector->min_vfreq = connector->display_info.monitor_range.min_vfreq; - amdgpu_dm_connector->max_vfreq = connector->display_info.monitor_range.max_vfreq; - if (amdgpu_dm_connector->max_vfreq - amdgpu_dm_connector->min_vfreq > 10) - freesync_capable = true; + if (amdgpu_dm_connector->dc_link && + amdgpu_dm_connector->dc_link->dpcd_caps.allow_invalid_MSA_timing_param) { + amdgpu_dm_connector->min_vfreq = connector->display_info.monitor_range.min_vfreq; + amdgpu_dm_connector->max_vfreq = connector->display_info.monitor_range.max_vfreq; + if (amdgpu_dm_connector->max_vfreq - amdgpu_dm_connector->min_vfreq > 10) + freesync_capable = true; + } + parse_amd_vsdb(amdgpu_dm_connector, edid, &vsdb_info); if (vsdb_info.replay_mode) {