From da3b36a23bb72e9742bf2f1b3e5da9615480c789 Mon Sep 17 00:00:00 2001 From: Jane Jian Date: Tue, 23 Nov 2021 19:19:40 +0800 Subject: drm/amdgpu/sriov/vcn: add new vcn ip revision check case for SIENNA_CICHLID [WHY] for sriov odd# vf will modify vcn0 engine ip revision(due to multimedia bandwidth feature), which will be mismatched with original vcn0 revision [HOW] add new version check for vcn0 disabled revision(3, 0, 192), typically modified under sriov mode Signed-off-by: Jane Jian Reviewed-by: Guchun Chen Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 1 + drivers/gpu/drm/amd/amdgpu/nv.c | 1 + 3 files changed, 3 insertions(+) (limited to 'drivers/gpu/drm/amd/amdgpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c index 503995c7ff6c..f6fae79203ee 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c @@ -918,6 +918,7 @@ static int amdgpu_discovery_set_mm_ip_blocks(struct amdgpu_device *adev) case IP_VERSION(3, 0, 64): case IP_VERSION(3, 1, 1): case IP_VERSION(3, 0, 2): + case IP_VERSION(3, 0, 192): amdgpu_device_ip_block_add(adev, &vcn_v3_0_ip_block); if (!amdgpu_sriov_vf(adev)) amdgpu_device_ip_block_add(adev, &jpeg_v3_0_ip_block); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c index 4f7c70845785..585961c2f5f2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c @@ -135,6 +135,7 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev) break; case IP_VERSION(3, 0, 0): case IP_VERSION(3, 0, 64): + case IP_VERSION(3, 0, 192): if (adev->ip_versions[GC_HWIP][0] == IP_VERSION(10, 3, 0)) fw_name = FIRMWARE_SIENNA_CICHLID; else diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c index a6659d9ecdd2..2ec1ffb36b1f 100644 --- a/drivers/gpu/drm/amd/amdgpu/nv.c +++ b/drivers/gpu/drm/amd/amdgpu/nv.c @@ -183,6 +183,7 @@ static int nv_query_video_codecs(struct amdgpu_device *adev, bool encode, switch (adev->ip_versions[UVD_HWIP][0]) { case IP_VERSION(3, 0, 0): case IP_VERSION(3, 0, 64): + case IP_VERSION(3, 0, 192): if (amdgpu_sriov_vf(adev)) { if (encode) *codecs = &sriov_sc_video_codecs_encode; -- cgit v1.2.3 From 3e467e478ed3a9701bb588d648d6e0ccb82ced09 Mon Sep 17 00:00:00 2001 From: Flora Cui Date: Wed, 24 Nov 2021 10:33:38 +0800 Subject: drm/amdgpu: cancel the correct hrtimer on exit Signed-off-by: Flora Cui Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c index ce982afeff91..ac9a8cd21c4b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vkms.c @@ -504,8 +504,8 @@ static int amdgpu_vkms_sw_fini(void *handle) int i = 0; for (i = 0; i < adev->mode_info.num_crtc; i++) - if (adev->mode_info.crtcs[i]) - hrtimer_cancel(&adev->mode_info.crtcs[i]->vblank_timer); + if (adev->amdgpu_vkms_output[i].vblank_hrtimer.function) + hrtimer_cancel(&adev->amdgpu_vkms_output[i].vblank_hrtimer); kfree(adev->mode_info.bios_hardcoded_edid); kfree(adev->amdgpu_vkms_output); -- cgit v1.2.3 From 1053b9c948e614473819a1a5bcaff6d44e680dcf Mon Sep 17 00:00:00 2001 From: Flora Cui Date: Thu, 18 Nov 2021 16:25:19 +0800 Subject: drm/amdgpu: check atomic flag to differeniate with legacy path since vkms support atomic KMS interface Signed-off-by: Flora Cui Reviewed-by: Guchun Chen Acked-by: Alex Deucher Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index d94fa748e6bb..4b7a69ef721f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -3833,7 +3833,7 @@ void amdgpu_device_fini_hw(struct amdgpu_device *adev) /* disable all interrupts */ amdgpu_irq_disable_all(adev); if (adev->mode_info.mode_config_initialized){ - if (!amdgpu_device_has_dc_support(adev)) + if (!drm_drv_uses_atomic_modeset(adev_to_drm(adev))) drm_helper_force_disable_all(adev_to_drm(adev)); else drm_atomic_helper_shutdown(adev_to_drm(adev)); @@ -5129,7 +5129,7 @@ skip_hw_reset: drm_sched_start(&ring->sched, !tmp_adev->asic_reset_res); } - if (!amdgpu_device_has_dc_support(tmp_adev) && !job_signaled) { + if (!drm_drv_uses_atomic_modeset(adev_to_drm(tmp_adev)) && !job_signaled) { drm_helper_resume_force_mode(adev_to_drm(tmp_adev)); } -- cgit v1.2.3 From 7551f70ab93d0f3371b28e996f7583e3be1d9a72 Mon Sep 17 00:00:00 2001 From: Guchun Chen Date: Fri, 26 Nov 2021 13:06:15 +0800 Subject: drm/amdgpu: fix the missed handling for SDMA2 and SDMA3 There is no base reg offset or ip_version set for SDMA2 and SDMA3 on SIENNA_CICHLID, so add them. Signed-off-by: Guchun Chen Reviewed-by: Kevin Wang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'drivers/gpu/drm/amd/amdgpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c index f6fae79203ee..ea00090b3fb3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c @@ -157,6 +157,8 @@ static int hw_id_map[MAX_HWIP] = { [HDP_HWIP] = HDP_HWID, [SDMA0_HWIP] = SDMA0_HWID, [SDMA1_HWIP] = SDMA1_HWID, + [SDMA2_HWIP] = SDMA2_HWID, + [SDMA3_HWIP] = SDMA3_HWID, [MMHUB_HWIP] = MMHUB_HWID, [ATHUB_HWIP] = ATHUB_HWID, [NBIO_HWIP] = NBIF_HWID, -- cgit v1.2.3 From e0570f0b6e2e88be7ef99d1194b153cb054a2107 Mon Sep 17 00:00:00 2001 From: Lijo Lazar Date: Mon, 29 Nov 2021 15:57:44 +0800 Subject: drm/amdgpu: Don't halt RLC on GFX suspend On aldebaran, RLC also controls GFXCLK. Skip halting RLC during GFX IP suspend and keep it running till PMFW disables all DPMs. [ 578.019986] amdgpu 0000:23:00.0: amdgpu: GPU reset begin! [ 583.245566] amdgpu 0000:23:00.0: amdgpu: Failed to disable smu features. [ 583.245621] amdgpu 0000:23:00.0: amdgpu: Fail to disable dpm features! [ 583.245639] [drm:amdgpu_device_ip_suspend_phase2 [amdgpu]] *ERROR* suspend of IP block failed -62 [ 583.248504] [drm] free PSP TMR buffer Signed-off-by: Lijo Lazar Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu') diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 34478bcc4d09..b305fd39874f 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -4060,9 +4060,10 @@ static int gfx_v9_0_hw_fini(void *handle) gfx_v9_0_cp_enable(adev, false); - /* Skip suspend with A+A reset */ - if (adev->gmc.xgmi.connected_to_cpu && amdgpu_in_reset(adev)) { - dev_dbg(adev->dev, "Device in reset. Skipping RLC halt\n"); + /* Skip stopping RLC with A+A reset or when RLC controls GFX clock */ + if ((adev->gmc.xgmi.connected_to_cpu && amdgpu_in_reset(adev)) || + (adev->ip_versions[GC_HWIP][0] >= IP_VERSION(9, 4, 2))) { + dev_dbg(adev->dev, "Skipping RLC halt\n"); return 0; } -- cgit v1.2.3 From 494f2e42ce4a9ddffb5d8c5b2db816425ef90397 Mon Sep 17 00:00:00 2001 From: Philip Yang Date: Fri, 26 Nov 2021 18:43:09 -0500 Subject: drm/amdkfd: fix double free mem structure drm_gem_object_put calls release_notify callback to free the mem structure and unreserve_mem_limit, move it down after the last access of mem and make it conditional call. Signed-off-by: Philip Yang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index 71a6a9ef54ac..6348559608ce 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -1396,7 +1396,7 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu( struct sg_table *sg = NULL; uint64_t user_addr = 0; struct amdgpu_bo *bo; - struct drm_gem_object *gobj; + struct drm_gem_object *gobj = NULL; u32 domain, alloc_domain; u64 alloc_flags; int ret; @@ -1506,14 +1506,16 @@ allocate_init_user_pages_failed: remove_kgd_mem_from_kfd_bo_list(*mem, avm->process_info); drm_vma_node_revoke(&gobj->vma_node, drm_priv); err_node_allow: - drm_gem_object_put(gobj); /* Don't unreserve system mem limit twice */ goto err_reserve_limit; err_bo_create: unreserve_mem_limit(adev, size, alloc_domain, !!sg); err_reserve_limit: mutex_destroy(&(*mem)->lock); - kfree(*mem); + if (gobj) + drm_gem_object_put(gobj); + else + kfree(*mem); err: if (sg) { sg_free_table(sg); -- cgit v1.2.3 From 428890a3fec131521cc59aac0d3c48bde9d76b7b Mon Sep 17 00:00:00 2001 From: shaoyunl Date: Mon, 29 Nov 2021 21:29:05 -0500 Subject: drm/amdgpu: adjust the kfd reset sequence in reset sriov function This change revert previous commits: 9f4f2c1a3524 ("drm/amd/amdgpu: fix the kfd pre_reset sequence in sriov") 271fd38ce56d ("drm/amdgpu: move kfd post_reset out of reset_sriov function") This change moves the amdgpu_amdkfd_pre_reset to an earlier place in amdgpu_device_reset_sriov, presumably to address the sequence issue that the first patch was originally meant to fix. Some register access(GRBM_GFX_CNTL) only be allowed on full access mode. Move kfd_pre_reset and kfd_post_reset back inside reset_sriov function. Fixes: 9f4f2c1a3524 ("drm/amd/amdgpu: fix the kfd pre_reset sequence in sriov") Fixes: 271fd38ce56d ("drm/amdgpu: move kfd post_reset out of reset_sriov function") Signed-off-by: shaoyunl Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 4b7a69ef721f..1e651b959141 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -4289,6 +4289,8 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev, { int r; + amdgpu_amdkfd_pre_reset(adev); + if (from_hypervisor) r = amdgpu_virt_request_full_gpu(adev, true); else @@ -4316,6 +4318,7 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev, amdgpu_irq_gpu_reset_resume_helper(adev); r = amdgpu_ib_ring_tests(adev); + amdgpu_amdkfd_post_reset(adev); error: if (!r && adev->virt.gim_feature & AMDGIM_FEATURE_GIM_FLR_VRAMLOST) { @@ -5030,7 +5033,8 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, cancel_delayed_work_sync(&tmp_adev->delayed_init_work); - amdgpu_amdkfd_pre_reset(tmp_adev); + if (!amdgpu_sriov_vf(tmp_adev)) + amdgpu_amdkfd_pre_reset(tmp_adev); /* * Mark these ASICs to be reseted as untracked first @@ -5148,9 +5152,9 @@ skip_hw_reset: skip_sched_resume: list_for_each_entry(tmp_adev, device_list_handle, reset_list) { - /* unlock kfd */ - if (!need_emergency_restart) - amdgpu_amdkfd_post_reset(tmp_adev); + /* unlock kfd: SRIOV would do it separately */ + if (!need_emergency_restart && !amdgpu_sriov_vf(tmp_adev)) + amdgpu_amdkfd_post_reset(tmp_adev); /* kfd_post_reset will do nothing if kfd device is not initialized, * need to bring up kfd here if it's not be initialized before -- cgit v1.2.3