From 31d5c52346cce168022b69b962023186b03eb43c Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 14 Mar 2022 14:38:19 -0400 Subject: drm/amdgpu: make amdgpu_display_framebuffer_init() static It's not used outside of amdgpu_display.c. Reviewed-by: Guchun Chen Reviewed-by: Aurabindo Pillai Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_display.c | 13 +++++++++---- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 4 ---- 2 files changed, 9 insertions(+), 8 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c index fae5c1debfad..b2bcc8b4eb82 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c @@ -41,6 +41,11 @@ #include #include +static int amdgpu_display_framebuffer_init(struct drm_device *dev, + struct amdgpu_framebuffer *rfb, + const struct drm_mode_fb_cmd2 *mode_cmd, + struct drm_gem_object *obj); + static void amdgpu_display_flip_callback(struct dma_fence *f, struct dma_fence_cb *cb) { @@ -1099,10 +1104,10 @@ err: return ret; } -int amdgpu_display_framebuffer_init(struct drm_device *dev, - struct amdgpu_framebuffer *rfb, - const struct drm_mode_fb_cmd2 *mode_cmd, - struct drm_gem_object *obj) +static int amdgpu_display_framebuffer_init(struct drm_device *dev, + struct amdgpu_framebuffer *rfb, + const struct drm_mode_fb_cmd2 *mode_cmd, + struct drm_gem_object *obj) { struct amdgpu_device *adev = drm_to_adev(dev); int ret, i; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index a546cb3cfa18..db9fe00486bc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -600,10 +600,6 @@ int amdgpu_display_gem_fb_verify_and_init( struct drm_device *dev, struct amdgpu_framebuffer *rfb, struct drm_file *file_priv, const struct drm_mode_fb_cmd2 *mode_cmd, struct drm_gem_object *obj); -int amdgpu_display_framebuffer_init(struct drm_device *dev, - struct amdgpu_framebuffer *rfb, - const struct drm_mode_fb_cmd2 *mode_cmd, - struct drm_gem_object *obj); int amdgpufb_remove(struct drm_device *dev, struct drm_framebuffer *fb); -- cgit v1.2.3 From 60da2f7440f22dba48944602c2a0373068f13880 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 14 Mar 2022 14:40:33 -0400 Subject: drm/amdgpu: drop amdgpu_display_gem_fb_init() Unused. Reviewed-by: Guchun Chen Reviewed-by: Aurabindo Pillai Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_display.c | 25 ------------------------- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 4 ---- 2 files changed, 29 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c index b2bcc8b4eb82..8eeaf6f12d35 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c @@ -1044,31 +1044,6 @@ static int amdgpu_display_get_fb_info(const struct amdgpu_framebuffer *amdgpu_fb return r; } -int amdgpu_display_gem_fb_init(struct drm_device *dev, - struct amdgpu_framebuffer *rfb, - const struct drm_mode_fb_cmd2 *mode_cmd, - struct drm_gem_object *obj) -{ - int ret; - - rfb->base.obj[0] = obj; - drm_helper_mode_fill_fb_struct(dev, &rfb->base, mode_cmd); - - ret = amdgpu_display_framebuffer_init(dev, rfb, mode_cmd, obj); - if (ret) - goto err; - - ret = drm_framebuffer_init(dev, &rfb->base, &amdgpu_fb_funcs); - if (ret) - goto err; - - return 0; -err: - drm_dbg_kms(dev, "Failed to init gem fb: %d\n", ret); - rfb->base.obj[0] = NULL; - return ret; -} - int amdgpu_display_gem_fb_verify_and_init( struct drm_device *dev, struct amdgpu_framebuffer *rfb, struct drm_file *file_priv, const struct drm_mode_fb_cmd2 *mode_cmd, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index db9fe00486bc..d637f3f2357c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -592,10 +592,6 @@ int amdgpu_display_get_crtc_scanoutpos(struct drm_device *dev, int *hpos, ktime_t *stime, ktime_t *etime, const struct drm_display_mode *mode); -int amdgpu_display_gem_fb_init(struct drm_device *dev, - struct amdgpu_framebuffer *rfb, - const struct drm_mode_fb_cmd2 *mode_cmd, - struct drm_gem_object *obj); int amdgpu_display_gem_fb_verify_and_init( struct drm_device *dev, struct amdgpu_framebuffer *rfb, struct drm_file *file_priv, const struct drm_mode_fb_cmd2 *mode_cmd, -- cgit v1.2.3 From c5b266810c5105b6c115de6236247049c23502db Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 14 Mar 2022 14:42:03 -0400 Subject: drm/amdgpu: make amdgpu_display_gem_fb_verify_and_init() static Unused outside of amdgpu_display.c. Reviewed-by: Guchun Chen Reviewed-by: Aurabindo Pillai Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_display.c | 9 +++++---- drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h | 5 ----- 2 files changed, 5 insertions(+), 9 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c index 8eeaf6f12d35..ee21b242ccf6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c @@ -1044,10 +1044,11 @@ static int amdgpu_display_get_fb_info(const struct amdgpu_framebuffer *amdgpu_fb return r; } -int amdgpu_display_gem_fb_verify_and_init( - struct drm_device *dev, struct amdgpu_framebuffer *rfb, - struct drm_file *file_priv, const struct drm_mode_fb_cmd2 *mode_cmd, - struct drm_gem_object *obj) +static int amdgpu_display_gem_fb_verify_and_init(struct drm_device *dev, + struct amdgpu_framebuffer *rfb, + struct drm_file *file_priv, + const struct drm_mode_fb_cmd2 *mode_cmd, + struct drm_gem_object *obj) { int ret; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h index d637f3f2357c..e8da738b309e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h @@ -592,11 +592,6 @@ int amdgpu_display_get_crtc_scanoutpos(struct drm_device *dev, int *hpos, ktime_t *stime, ktime_t *etime, const struct drm_display_mode *mode); -int amdgpu_display_gem_fb_verify_and_init( - struct drm_device *dev, struct amdgpu_framebuffer *rfb, - struct drm_file *file_priv, const struct drm_mode_fb_cmd2 *mode_cmd, - struct drm_gem_object *obj); - int amdgpufb_remove(struct drm_device *dev, struct drm_framebuffer *fb); void amdgpu_enc_destroy(struct drm_encoder *encoder); -- cgit v1.2.3 From eed41975307a365d1c0f6ed5c388b130ffdd2664 Mon Sep 17 00:00:00 2001 From: Tao Zhou Date: Tue, 15 Mar 2022 17:07:21 +0800 Subject: drm/amdkfd: refine event_interrupt_poison_consumption Combine reading and setting poison flag as one atomic operation and add print message for the function. Signed-off-by: Tao Zhou Reviewed-by: Hawking Zhang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c index 7eedbcd14828..a992798ff8b6 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c @@ -93,20 +93,19 @@ enum SQ_INTERRUPT_ERROR_TYPE { static void event_interrupt_poison_consumption(struct kfd_dev *dev, uint16_t pasid, uint16_t source_id) { - int ret = -EINVAL; + int old_poison, ret = -EINVAL; struct kfd_process *p = kfd_lookup_process_by_pasid(pasid); if (!p) return; /* all queues of a process will be unmapped in one time */ - if (atomic_read(&p->poison)) { - kfd_unref_process(p); + old_poison = atomic_cmpxchg(&p->poison, 0, 1); + kfd_unref_process(p); + if (old_poison) return; - } - atomic_set(&p->poison, 1); - kfd_unref_process(p); + pr_warn("RAS poison consumption handling\n"); switch (source_id) { case SOC15_INTSRC_SQ_INTERRUPT_MSG: -- cgit v1.2.3 From 9d8a8d78d95261241fdc009ff5b44cfa9f78c8e7 Mon Sep 17 00:00:00 2001 From: Tao Zhou Date: Wed, 16 Mar 2022 12:03:51 +0800 Subject: drm/amdkfd: replace source_id with client_id for RAS poison consumption Client ID is more accruate here and we can deal with more different cases with client ID. Signed-off-by: Tao Zhou Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c | 23 ++++++++++++++++------- 1 file changed, 16 insertions(+), 7 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c index a992798ff8b6..7db2421a3340 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c @@ -91,7 +91,7 @@ enum SQ_INTERRUPT_ERROR_TYPE { #define KFD_SQ_INT_DATA__ERR_TYPE__SHIFT 20 static void event_interrupt_poison_consumption(struct kfd_dev *dev, - uint16_t pasid, uint16_t source_id) + uint16_t pasid, uint16_t client_id) { int old_poison, ret = -EINVAL; struct kfd_process *p = kfd_lookup_process_by_pasid(pasid); @@ -105,13 +105,22 @@ static void event_interrupt_poison_consumption(struct kfd_dev *dev, if (old_poison) return; - pr_warn("RAS poison consumption handling\n"); + pr_warn("RAS poison consumption handling: client id %d\n", client_id); - switch (source_id) { - case SOC15_INTSRC_SQ_INTERRUPT_MSG: + switch (client_id) { + case SOC15_IH_CLIENTID_SE0SH: + case SOC15_IH_CLIENTID_SE1SH: + case SOC15_IH_CLIENTID_SE2SH: + case SOC15_IH_CLIENTID_SE3SH: + case SOC15_IH_CLIENTID_UTCL2: ret = kfd_dqm_evict_pasid(dev->dqm, pasid); break; - case SOC15_INTSRC_SDMA_ECC: + case SOC15_IH_CLIENTID_SDMA0: + case SOC15_IH_CLIENTID_SDMA1: + case SOC15_IH_CLIENTID_SDMA2: + case SOC15_IH_CLIENTID_SDMA3: + case SOC15_IH_CLIENTID_SDMA4: + break; default: break; } @@ -269,7 +278,7 @@ static void event_interrupt_wq_v9(struct kfd_dev *dev, sq_intr_err); if (sq_intr_err != SQ_INTERRUPT_ERROR_TYPE_ILLEGAL_INST && sq_intr_err != SQ_INTERRUPT_ERROR_TYPE_MEMVIOL) { - event_interrupt_poison_consumption(dev, pasid, source_id); + event_interrupt_poison_consumption(dev, pasid, client_id); return; } break; @@ -290,7 +299,7 @@ static void event_interrupt_wq_v9(struct kfd_dev *dev, if (source_id == SOC15_INTSRC_SDMA_TRAP) { kfd_signal_event_interrupt(pasid, context_id0 & 0xfffffff, 28); } else if (source_id == SOC15_INTSRC_SDMA_ECC) { - event_interrupt_poison_consumption(dev, pasid, source_id); + event_interrupt_poison_consumption(dev, pasid, client_id); return; } } else if (client_id == SOC15_IH_CLIENTID_VMC || -- cgit v1.2.3 From 6475ae2b742876aa9b2a0aff7ba60f5c81917614 Mon Sep 17 00:00:00 2001 From: Tao Zhou Date: Tue, 15 Mar 2022 17:48:18 +0800 Subject: drm/amdgpu: add UTCL2 RAS poison query for Aldebaran (v2) Add help functions to query and reset RAS UTCL2 poison status. v2: implement it on amdgpu side and kfd only calls it. Signed-off-by: Tao Zhou Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 8 ++++++++ drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h | 1 + drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c | 14 ++++++++++++++ 4 files changed, 24 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index 6ca1db3c243f..c18c4be1e4ac 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -724,3 +724,11 @@ void amdgpu_amdkfd_ras_poison_consumption_handler(struct amdgpu_device *adev, bo else if (reset) amdgpu_amdkfd_gpu_reset(adev); } + +bool amdgpu_amdkfd_ras_query_utcl2_poison_status(struct amdgpu_device *adev) +{ + if (adev->gfx.ras->query_utcl2_poison_status) + return adev->gfx.ras->query_utcl2_poison_status(adev); + else + return false; +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index 4cb14c2fe53f..0838926a8ef0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -301,6 +301,7 @@ void amdgpu_amdkfd_ras_poison_consumption_handler(struct amdgpu_device *adev, bool amdgpu_amdkfd_bo_mapped_to_dev(struct amdgpu_device *adev, struct kgd_mem *mem); void amdgpu_amdkfd_block_mmu_notifications(void *p); int amdgpu_amdkfd_criu_resume(void *p); +bool amdgpu_amdkfd_ras_query_utcl2_poison_status(struct amdgpu_device *adev); #if IS_ENABLED(CONFIG_HSA_AMD) void amdgpu_amdkfd_gpuvm_init_mem_limits(void); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h index dcb3c7871c73..5ed9b8a4c571 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h @@ -202,6 +202,7 @@ struct amdgpu_cu_info { struct amdgpu_gfx_ras { struct amdgpu_ras_block_object ras_block; void (*enable_watchdog_timer)(struct amdgpu_device *adev); + bool (*query_utcl2_poison_status)(struct amdgpu_device *adev); }; struct amdgpu_gfx_funcs { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c index 7653ebd0e67b..3a797424579c 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c @@ -1930,6 +1930,19 @@ static void gfx_v9_4_2_reset_sq_timeout_status(struct amdgpu_device *adev) mutex_unlock(&adev->grbm_idx_mutex); } +static bool gfx_v9_4_2_query_uctl2_poison_status(struct amdgpu_device *adev) +{ + u32 status = 0; + struct amdgpu_vmhub *hub; + + hub = &adev->vmhub[AMDGPU_GFXHUB_0]; + status = RREG32(hub->vm_l2_pro_fault_status); + /* reset page fault status */ + WREG32_P(hub->vm_l2_pro_fault_cntl, 1, ~1); + + return REG_GET_FIELD(status, VM_L2_PROTECTION_FAULT_STATUS, FED); +} + struct amdgpu_ras_block_hw_ops gfx_v9_4_2_ras_ops = { .ras_error_inject = &gfx_v9_4_2_ras_error_inject, .query_ras_error_count = &gfx_v9_4_2_query_ras_error_count, @@ -1943,4 +1956,5 @@ struct amdgpu_gfx_ras gfx_v9_4_2_ras = { .hw_ops = &gfx_v9_4_2_ras_ops, }, .enable_watchdog_timer = &gfx_v9_4_2_enable_watchdog_timer, + .query_utcl2_poison_status = gfx_v9_4_2_query_uctl2_poison_status, }; -- cgit v1.2.3 From 1990e29b1900758f596434204d4067955f6e904e Mon Sep 17 00:00:00 2001 From: Tao Zhou Date: Wed, 16 Mar 2022 14:38:12 +0800 Subject: drm/amdkfd: add RAS poison consumption handling for UTCL2 (v2) Do RAS page retirement and use gpu reset as fallback in UTCL2 fault handler. v2: replace vm fault event with posion consumed event in UTCL2 poison consumption. Signed-off-by: Tao Zhou Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c index 7db2421a3340..56902b5bb7b6 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c @@ -308,6 +308,12 @@ static void event_interrupt_wq_v9(struct kfd_dev *dev, struct kfd_vm_fault_info info = {0}; uint16_t ring_id = SOC15_RING_ID_FROM_IH_ENTRY(ih_ring_entry); + if (client_id == SOC15_IH_CLIENTID_UTCL2 && + amdgpu_amdkfd_ras_query_utcl2_poison_status(dev->adev)) { + event_interrupt_poison_consumption(dev, pasid, client_id); + return; + } + info.vmid = vmid; info.mc_id = client_id; info.page_addr = ih_ring_entry[4] | -- cgit v1.2.3 From 2f78f0d3e3a89e76a7fe184fdef19efd85063f61 Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Tue, 15 Mar 2022 20:19:59 +0000 Subject: drm/amdgpu: Fix spelling mistake "regiser" -> "register" There is a spelling mistake in a dev_error error message. Fix it. Signed-off-by: Colin Ian King Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index a025f080aa6a..9aa355a5ac3c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -919,7 +919,7 @@ static u32 amdgpu_virt_rlcg_reg_rw(struct amdgpu_device *adev, u32 offset, u32 v "wrong operation type, rlcg failed to program reg: 0x%05x\n", offset); } else if (tmp & AMDGPU_RLCG_REG_NOT_IN_RANGE) { dev_err(adev->dev, - "regiser is not in range, rlcg failed to program reg: 0x%05x\n", offset); + "register is not in range, rlcg failed to program reg: 0x%05x\n", offset); } else { dev_err(adev->dev, "unknown error type, rlcg failed to program reg: 0x%05x\n", offset); -- cgit v1.2.3 From f3fa490960e8acf089ace0301a7ccb6d44ec9641 Mon Sep 17 00:00:00 2001 From: Guchun Chen Date: Wed, 16 Mar 2022 14:51:31 +0800 Subject: drm/amdgpu: drop redundant check of harvest info Harvest bit setting in IP data structure promises this, so no need to set it explicitly. Signed-off-by: Guchun Chen Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c | 7 ------- 1 file changed, 7 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c index e4fcbb385a62..5a182288391b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c @@ -1150,13 +1150,6 @@ void amdgpu_discovery_harvest_ip(struct amdgpu_device *adev) adev->harvest_ip_mask |= AMD_HARVEST_IP_VCN_MASK; adev->harvest_ip_mask |= AMD_HARVEST_IP_JPEG_MASK; } - if ((adev->pdev->device == 0x731E && - (adev->pdev->revision == 0xC6 || adev->pdev->revision == 0xC7)) || - (adev->pdev->device == 0x7340 && adev->pdev->revision == 0xC9) || - (adev->pdev->device == 0x7360 && adev->pdev->revision == 0xC7)) { - adev->harvest_ip_mask |= AMD_HARVEST_IP_VCN_MASK; - adev->harvest_ip_mask |= AMD_HARVEST_IP_JPEG_MASK; - } } union gc_info { -- cgit v1.2.3 From 4268d081619a55019e44bf0a62222d47e1e09470 Mon Sep 17 00:00:00 2001 From: Ahmad Othman Date: Fri, 4 Mar 2022 11:56:19 -0500 Subject: drm/amd/display: Fix HDCP SEND AKI INIT error [why] HDCP sends AKI INIT error in case of multiple display on dock [how] Add new checks and method to handle display adjustment for multiple display cases Reviewed-by: Wenjing Liu Acked-by: Alex Hung Signed-off-by: Ahmad Othman Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c | 38 +++++++++++++++++++++- drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h | 8 +++++ drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h | 2 +- 3 files changed, 46 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c index 3e81850a7ffe..5e01c6e24cbc 100644 --- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c +++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c @@ -251,6 +251,33 @@ out: return status; } +static enum mod_hdcp_status update_display_adjustments(struct mod_hdcp *hdcp, + struct mod_hdcp_display *display, + struct mod_hdcp_display_adjustment *adj) +{ + enum mod_hdcp_status status = MOD_HDCP_STATUS_NOT_IMPLEMENTED; + + if (is_in_authenticated_states(hdcp) && + is_dp_mst_hdcp(hdcp) && + display->adjust.disable == true && + adj->disable == false) { + display->adjust.disable = false; + if (is_hdcp1(hdcp)) + status = mod_hdcp_hdcp1_enable_dp_stream_encryption(hdcp); + else if (is_hdcp2(hdcp)) + status = mod_hdcp_hdcp2_enable_dp_stream_encryption(hdcp); + + if (status != MOD_HDCP_STATUS_SUCCESS) + display->adjust.disable = true; + } + + if (status == MOD_HDCP_STATUS_SUCCESS && + memcmp(adj, &display->adjust, + sizeof(struct mod_hdcp_display_adjustment)) != 0) + status = MOD_HDCP_STATUS_NOT_IMPLEMENTED; + + return status; +} /* * Implementation of functions in mod_hdcp.h */ @@ -391,7 +418,7 @@ out: return status; } -enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, +enum mod_hdcp_status mod_hdcp_update_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_link_adjustment *link_adjust, struct mod_hdcp_display_adjustment *display_adjust, @@ -419,6 +446,15 @@ enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, goto out; } + if (memcmp(link_adjust, &hdcp->connection.link.adjust, + sizeof(struct mod_hdcp_link_adjustment)) == 0 && + memcmp(display_adjust, &display->adjust, + sizeof(struct mod_hdcp_display_adjustment)) != 0) { + status = update_display_adjustments(hdcp, display, display_adjust); + if (status != MOD_HDCP_STATUS_NOT_IMPLEMENTED) + goto out; + } + /* stop current authentication */ status = reset_authentication(hdcp, output); if (status != MOD_HDCP_STATUS_SUCCESS) diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h index 8502263d2968..4e7021c3c845 100644 --- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h +++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h @@ -445,6 +445,14 @@ static inline uint8_t is_in_hdcp2_dp_states(struct mod_hdcp *hdcp) current_state(hdcp) <= HDCP2_DP_STATE_END); } +static inline uint8_t is_in_authenticated_states(struct mod_hdcp *hdcp) +{ + return (current_state(hdcp) == D1_A4_AUTHENTICATED || + current_state(hdcp) == H1_A45_AUTHENTICATED || + current_state(hdcp) == D2_A5_AUTHENTICATED || + current_state(hdcp) == H2_A5_AUTHENTICATED); +} + static inline uint8_t is_hdcp1(struct mod_hdcp *hdcp) { return (is_in_hdcp1_states(hdcp) || is_in_hdcp1_dp_states(hdcp)); diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h b/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h index f7420c3f5672..3348bb97ef81 100644 --- a/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h +++ b/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h @@ -294,7 +294,7 @@ enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_output *output); /* called per display to apply new authentication adjustment */ -enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, +enum mod_hdcp_status mod_hdcp_update_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_link_adjustment *link_adjust, struct mod_hdcp_display_adjustment *display_adjust, -- cgit v1.2.3 From ed94aca6dbca2519cb71a73c2d276c97fe857596 Mon Sep 17 00:00:00 2001 From: Tao Zhou Date: Mon, 21 Mar 2022 15:45:31 +0800 Subject: drm/amdkfd: print unmap queue status for RAS poison consumption (v3) Print the status out when it passes, and also tell user gpu reset is triggered when we fall back to legacy way. v2: make the message more explicit. v3: change succeeds to succeeded. replace pr_warn with dev_warn. Signed-off-by: Tao Zhou Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c index 56902b5bb7b6..03c29bdd89a1 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c @@ -105,8 +105,6 @@ static void event_interrupt_poison_consumption(struct kfd_dev *dev, if (old_poison) return; - pr_warn("RAS poison consumption handling: client id %d\n", client_id); - switch (client_id) { case SOC15_IH_CLIENTID_SE0SH: case SOC15_IH_CLIENTID_SE1SH: @@ -130,10 +128,17 @@ static void event_interrupt_poison_consumption(struct kfd_dev *dev, /* resetting queue passes, do page retirement without gpu reset * resetting queue fails, fallback to gpu reset solution */ - if (!ret) + if (!ret) { + dev_warn(dev->adev->dev, + "RAS poison consumption, unmap queue flow succeeded: client id %d\n", + client_id); amdgpu_amdkfd_ras_poison_consumption_handler(dev->adev, false); - else + } else { + dev_warn(dev->adev->dev, + "RAS poison consumption, fall back to gpu reset flow: client id %d\n", + client_id); amdgpu_amdkfd_ras_poison_consumption_handler(dev->adev, true); + } } static bool event_interrupt_isr_v9(struct kfd_dev *dev, -- cgit v1.2.3 From 31f714b44959821f44fae105bd1d024b0f6b7ed5 Mon Sep 17 00:00:00 2001 From: Anthony Koo Date: Sun, 13 Mar 2022 14:06:05 -0400 Subject: drm/amd/display: [FW Promotion] Release 0.0.109.0 Reviewed-by: Aric Cyr Acked-by: Alex Hung Signed-off-by: Anthony Koo Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') 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 71214c7a60fc..ce773b56a778 100644 --- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h @@ -46,10 +46,10 @@ /* Firmware versioning. */ #ifdef DMUB_EXPOSE_VERSION -#define DMUB_FW_VERSION_GIT_HASH 0x929554ba +#define DMUB_FW_VERSION_GIT_HASH 0x51b95a35 #define DMUB_FW_VERSION_MAJOR 0 #define DMUB_FW_VERSION_MINOR 0 -#define DMUB_FW_VERSION_REVISION 108 +#define DMUB_FW_VERSION_REVISION 109 #define DMUB_FW_VERSION_TEST 0 #define DMUB_FW_VERSION_VBIOS 0 #define DMUB_FW_VERSION_HOTFIX 0 -- cgit v1.2.3 From 1918a0455d360457b2fbcda889cff19876d01b9e Mon Sep 17 00:00:00 2001 From: Aric Cyr Date: Sun, 13 Mar 2022 21:04:59 -0400 Subject: drm/amd/display: 3.2.178 This version brings along following fixes: - HDCP SEND AKI INIT error - fix audio format not updated after edid updated - Reduce stack size - FEC check in timing validation - Add fSMC_MSG_SetDtbClk support - Update VTEM Infopacket definition - [FW Promotion] Release 0.0.109.0 Acked-by: Alex Hung Signed-off-by: Aric Cyr 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(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 9f4d926d54e7..ced40fe218ac 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -47,7 +47,7 @@ struct aux_payload; struct set_config_cmd_payload; struct dmub_notification; -#define DC_VER "3.2.177" +#define DC_VER "3.2.178" #define MAX_SURFACES 3 #define MAX_PLANES 6 -- cgit v1.2.3 From 26f4712aedbdf4b9f5e3888a50a2a4b130ee4a9b Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Mon, 7 Mar 2022 14:47:59 -0100 Subject: drm/amd/display: move FPU related code from dcn31 to dml/dcn31 folder Creates FPU files in dml/dcn31 folder to centralize FPU operations from 3.1x drivers and moves all FPU-associated code from dcn31 driver to there. It includes the struct _vcs_dpi_ip_params_st and _vcs_dpi_soc_bounding_box_st and functions: - dcn31_calculate_wm_and_dlg_fp() - dcn31_update_bw_bounding_box() adding dc_assert_fp_enabled to them and drop DC_FP_START/END inside functions that was moved to dml folder, as required. Signed-off-by: Melissa Wen Reviewed-by: Alex Hung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn31/Makefile | 26 -- .../gpu/drm/amd/display/dc/dcn31/dcn31_resource.c | 355 +----------------- .../gpu/drm/amd/display/dc/dcn31/dcn31_resource.h | 4 +- drivers/gpu/drm/amd/display/dc/dml/Makefile | 2 + .../gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c | 406 +++++++++++++++++++++ .../gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h | 39 ++ 6 files changed, 451 insertions(+), 381 deletions(-) create mode 100644 drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c create mode 100644 drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/Makefile b/drivers/gpu/drm/amd/display/dc/dcn31/Makefile index d20e3b8ccc30..ec041e3cda30 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dcn31/Makefile @@ -15,32 +15,6 @@ DCN31 = dcn31_resource.o dcn31_hubbub.o dcn31_hwseq.o dcn31_init.o dcn31_hubp.o dcn31_apg.o dcn31_hpo_dp_stream_encoder.o dcn31_hpo_dp_link_encoder.o \ dcn31_afmt.o dcn31_vpg.o -ifdef CONFIG_X86 -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o := -msse -endif - -ifdef CONFIG_PPC64 -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o := -mhard-float -maltivec -endif - -ifdef CONFIG_CC_IS_GCC -ifeq ($(call cc-ifversion, -lt, 0701, y), y) -IS_OLD_GCC = 1 -endif -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o += -mhard-float -endif - -ifdef CONFIG_X86 -ifdef IS_OLD_GCC -# Stack alignment mismatch, proceed with caution. -# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3 -# (8B stack alignment). -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o += -mpreferred-stack-boundary=4 -else -CFLAGS_$(AMDDALPATH)/dc/dcn31/dcn31_resource.o += -msse2 -endif -endif - AMD_DAL_DCN31 = $(addprefix $(AMDDALPATH)/dc/dcn31/,$(DCN31)) AMD_DISPLAY_FILES += $(AMD_DAL_DCN31) diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c index 338235bcef4a..bf130b2435ab 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c @@ -65,6 +65,7 @@ #include "virtual/virtual_stream_encoder.h" #include "dce110/dce110_resource.h" #include "dml/display_mode_vba.h" +#include "dml/dcn31/dcn31_fpu.h" #include "dcn31/dcn31_dccg.h" #include "dcn10/dcn10_resource.h" #include "dcn31_panel_cntl.h" @@ -102,152 +103,6 @@ #define DC_LOGGER_INIT(logger) -#define DCN3_1_DEFAULT_DET_SIZE 384 - -struct _vcs_dpi_ip_params_st dcn3_1_ip = { - .gpuvm_enable = 1, - .gpuvm_max_page_table_levels = 1, - .hostvm_enable = 1, - .hostvm_max_page_table_levels = 2, - .rob_buffer_size_kbytes = 64, - .det_buffer_size_kbytes = DCN3_1_DEFAULT_DET_SIZE, - .config_return_buffer_size_in_kbytes = 1792, - .compressed_buffer_segment_size_in_kbytes = 64, - .meta_fifo_size_in_kentries = 32, - .zero_size_buffer_entries = 512, - .compbuf_reserved_space_64b = 256, - .compbuf_reserved_space_zs = 64, - .dpp_output_buffer_pixels = 2560, - .opp_output_buffer_lines = 1, - .pixel_chunk_size_kbytes = 8, - .meta_chunk_size_kbytes = 2, - .min_meta_chunk_size_bytes = 256, - .writeback_chunk_size_kbytes = 8, - .ptoi_supported = false, - .num_dsc = 3, - .maximum_dsc_bits_per_component = 10, - .dsc422_native_support = false, - .is_line_buffer_bpp_fixed = true, - .line_buffer_fixed_bpp = 48, - .line_buffer_size_bits = 789504, - .max_line_buffer_lines = 12, - .writeback_interface_buffer_size_kbytes = 90, - .max_num_dpp = 4, - .max_num_otg = 4, - .max_num_hdmi_frl_outputs = 1, - .max_num_wb = 1, - .max_dchub_pscl_bw_pix_per_clk = 4, - .max_pscl_lb_bw_pix_per_clk = 2, - .max_lb_vscl_bw_pix_per_clk = 4, - .max_vscl_hscl_bw_pix_per_clk = 4, - .max_hscl_ratio = 6, - .max_vscl_ratio = 6, - .max_hscl_taps = 8, - .max_vscl_taps = 8, - .dpte_buffer_size_in_pte_reqs_luma = 64, - .dpte_buffer_size_in_pte_reqs_chroma = 34, - .dispclk_ramp_margin_percent = 1, - .max_inter_dcn_tile_repeaters = 8, - .cursor_buffer_size = 16, - .cursor_chunk_size = 2, - .writeback_line_buffer_buffer_size = 0, - .writeback_min_hscl_ratio = 1, - .writeback_min_vscl_ratio = 1, - .writeback_max_hscl_ratio = 1, - .writeback_max_vscl_ratio = 1, - .writeback_max_hscl_taps = 1, - .writeback_max_vscl_taps = 1, - .dppclk_delay_subtotal = 46, - .dppclk_delay_scl = 50, - .dppclk_delay_scl_lb_only = 16, - .dppclk_delay_cnvc_formatter = 27, - .dppclk_delay_cnvc_cursor = 6, - .dispclk_delay_subtotal = 119, - .dynamic_metadata_vm_enabled = false, - .odm_combine_4to1_supported = false, - .dcc_supported = true, -}; - -struct _vcs_dpi_soc_bounding_box_st dcn3_1_soc = { - /*TODO: correct dispclk/dppclk voltage level determination*/ - .clock_limits = { - { - .state = 0, - .dispclk_mhz = 1200.0, - .dppclk_mhz = 1200.0, - .phyclk_mhz = 600.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 186.0, - .dtbclk_mhz = 625.0, - }, - { - .state = 1, - .dispclk_mhz = 1200.0, - .dppclk_mhz = 1200.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 209.0, - .dtbclk_mhz = 625.0, - }, - { - .state = 2, - .dispclk_mhz = 1200.0, - .dppclk_mhz = 1200.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 209.0, - .dtbclk_mhz = 625.0, - }, - { - .state = 3, - .dispclk_mhz = 1200.0, - .dppclk_mhz = 1200.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 371.0, - .dtbclk_mhz = 625.0, - }, - { - .state = 4, - .dispclk_mhz = 1200.0, - .dppclk_mhz = 1200.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 417.0, - .dtbclk_mhz = 625.0, - }, - }, - .num_states = 5, - .sr_exit_time_us = 9.0, - .sr_enter_plus_exit_time_us = 11.0, - .sr_exit_z8_time_us = 442.0, - .sr_enter_plus_exit_z8_time_us = 560.0, - .writeback_latency_us = 12.0, - .dram_channel_width_bytes = 4, - .round_trip_ping_latency_dcfclk_cycles = 106, - .urgent_latency_pixel_data_only_us = 4.0, - .urgent_latency_pixel_mixed_with_vm_data_us = 4.0, - .urgent_latency_vm_data_only_us = 4.0, - .urgent_out_of_order_return_per_channel_pixel_only_bytes = 4096, - .urgent_out_of_order_return_per_channel_pixel_and_vm_bytes = 4096, - .urgent_out_of_order_return_per_channel_vm_only_bytes = 4096, - .pct_ideal_sdp_bw_after_urgent = 80.0, - .pct_ideal_dram_sdp_bw_after_urgent_pixel_only = 65.0, - .pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm = 60.0, - .pct_ideal_dram_sdp_bw_after_urgent_vm_only = 30.0, - .max_avg_sdp_bw_use_normal_percent = 60.0, - .max_avg_dram_bw_use_normal_percent = 60.0, - .fabric_datapath_to_dcn_data_return_bytes = 32, - .return_bus_width_bytes = 64, - .downspread_percent = 0.38, - .dcn_downspread_percent = 0.5, - .gpuvm_min_page_size_bytes = 4096, - .hostvm_min_page_size_bytes = 4096, - .do_urgent_latency_adjustment = false, - .urgent_latency_adjustment_fabric_clock_component_us = 0, - .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, -}; - enum dcn31_clk_src_array_id { DCN31_CLK_SRC_PLL0, DCN31_CLK_SRC_PLL1, @@ -1869,143 +1724,6 @@ void dcn31_update_soc_for_wm_a(struct dc *dc, struct dc_state *context) } } -static void dcn31_calculate_wm_and_dlg_fp( - struct dc *dc, struct dc_state *context, - display_e2e_pipe_params_st *pipes, - int pipe_cnt, - int vlevel) -{ - int i, pipe_idx; - double dcfclk = context->bw_ctx.dml.vba.DCFCLKState[vlevel][context->bw_ctx.dml.vba.maxMpcComb]; - - if (context->bw_ctx.dml.soc.min_dcfclk > dcfclk) - dcfclk = context->bw_ctx.dml.soc.min_dcfclk; - - /* We don't recalculate clocks for 0 pipe configs, which can block - * S0i3 as high clocks will block low power states - * Override any clocks that can block S0i3 to min here - */ - if (pipe_cnt == 0) { - context->bw_ctx.bw.dcn.clk.dcfclk_khz = dcfclk; // always should be vlevel 0 - return; - } - - pipes[0].clks_cfg.voltage = vlevel; - pipes[0].clks_cfg.dcfclk_mhz = dcfclk; - pipes[0].clks_cfg.socclk_mhz = context->bw_ctx.dml.soc.clock_limits[vlevel].socclk_mhz; - -#if 0 // TODO - /* Set B: - * TODO - */ - if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B].valid) { - if (vlevel == 0) { - pipes[0].clks_cfg.voltage = 1; - pipes[0].clks_cfg.dcfclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dcfclk_mhz; - } - context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.pstate_latency_us; - context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.sr_enter_plus_exit_time_us; - context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.sr_exit_time_us; - } - context->bw_ctx.bw.dcn.watermarks.b.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_z8_ns = get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_z8_ns = get_wm_z8_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.b.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - - pipes[0].clks_cfg.voltage = vlevel; - pipes[0].clks_cfg.dcfclk_mhz = dcfclk; - - /* Set C: - * TODO - */ - if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].valid) { - context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.pstate_latency_us; - context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.sr_enter_plus_exit_time_us; - context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.sr_exit_time_us; - } - context->bw_ctx.bw.dcn.watermarks.c.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_z8_ns = get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_z8_ns = get_wm_z8_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.c.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - - /* Set D: - * TODO - */ - if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].valid) { - context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.pstate_latency_us; - context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.sr_enter_plus_exit_time_us; - context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.sr_exit_time_us; - } - context->bw_ctx.bw.dcn.watermarks.d.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_z8_ns = get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_z8_ns = get_wm_z8_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.d.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; -#endif - - /* Set A: - * All clocks min required - * - * Set A calculated last so that following calculations are based on Set A - */ - dc->res_pool->funcs->update_soc_for_wm_a(dc, context); - context->bw_ctx.bw.dcn.watermarks.a.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_z8_ns = get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_z8_ns = get_wm_z8_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.a.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - context->bw_ctx.bw.dcn.watermarks.a.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; - /* TODO: remove: */ - context->bw_ctx.bw.dcn.watermarks.b = context->bw_ctx.bw.dcn.watermarks.a; - context->bw_ctx.bw.dcn.watermarks.c = context->bw_ctx.bw.dcn.watermarks.a; - context->bw_ctx.bw.dcn.watermarks.d = context->bw_ctx.bw.dcn.watermarks.a; - /* end remove*/ - - for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) { - if (!context->res_ctx.pipe_ctx[i].stream) - continue; - - pipes[pipe_idx].clks_cfg.dispclk_mhz = get_dispclk_calculated(&context->bw_ctx.dml, pipes, pipe_cnt); - pipes[pipe_idx].clks_cfg.dppclk_mhz = get_dppclk_calculated(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx); - - if (dc->config.forced_clocks || dc->debug.max_disp_clk) { - pipes[pipe_idx].clks_cfg.dispclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dispclk_mhz; - pipes[pipe_idx].clks_cfg.dppclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dppclk_mhz; - } - if (dc->debug.min_disp_clk_khz > pipes[pipe_idx].clks_cfg.dispclk_mhz * 1000) - pipes[pipe_idx].clks_cfg.dispclk_mhz = dc->debug.min_disp_clk_khz / 1000.0; - if (dc->debug.min_dpp_clk_khz > pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000) - pipes[pipe_idx].clks_cfg.dppclk_mhz = dc->debug.min_dpp_clk_khz / 1000.0; - - pipe_idx++; - } - - DC_FP_START(); - dcn20_calculate_dlg_params(dc, context, pipes, pipe_cnt, vlevel); - DC_FP_END(); -} - void dcn31_calculate_wm_and_dlg( struct dc *dc, struct dc_state *context, display_e2e_pipe_params_st *pipes, @@ -2073,77 +1791,6 @@ static struct dc_cap_funcs cap_funcs = { .get_dcc_compression_cap = dcn20_get_dcc_compression_cap }; -void dcn31_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params) -{ - struct clk_limit_table *clk_table = &bw_params->clk_table; - struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES]; - unsigned int i, closest_clk_lvl; - int j; - - // Default clock levels are used for diags, which may lead to overclocking. - if (!IS_DIAG_DC(dc->ctx->dce_environment)) { - int max_dispclk_mhz = 0, max_dppclk_mhz = 0; - - dcn3_1_ip.max_num_otg = dc->res_pool->res_cap->num_timing_generator; - dcn3_1_ip.max_num_dpp = dc->res_pool->pipe_count; - dcn3_1_soc.num_chans = bw_params->num_channels; - - ASSERT(clk_table->num_entries); - - /* Prepass to find max clocks independent of voltage level. */ - for (i = 0; i < clk_table->num_entries; ++i) { - if (clk_table->entries[i].dispclk_mhz > max_dispclk_mhz) - max_dispclk_mhz = clk_table->entries[i].dispclk_mhz; - if (clk_table->entries[i].dppclk_mhz > max_dppclk_mhz) - max_dppclk_mhz = clk_table->entries[i].dppclk_mhz; - } - - for (i = 0; i < clk_table->num_entries; i++) { - /* loop backwards*/ - for (closest_clk_lvl = 0, j = dcn3_1_soc.num_states - 1; j >= 0; j--) { - if ((unsigned int) dcn3_1_soc.clock_limits[j].dcfclk_mhz <= clk_table->entries[i].dcfclk_mhz) { - closest_clk_lvl = j; - break; - } - } - - clock_limits[i].state = i; - - /* Clocks dependent on voltage level. */ - clock_limits[i].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz; - clock_limits[i].fabricclk_mhz = clk_table->entries[i].fclk_mhz; - clock_limits[i].socclk_mhz = clk_table->entries[i].socclk_mhz; - clock_limits[i].dram_speed_mts = clk_table->entries[i].memclk_mhz * 2 * clk_table->entries[i].wck_ratio; - - /* Clocks independent of voltage level. */ - clock_limits[i].dispclk_mhz = max_dispclk_mhz ? max_dispclk_mhz : - dcn3_1_soc.clock_limits[closest_clk_lvl].dispclk_mhz; - - clock_limits[i].dppclk_mhz = max_dppclk_mhz ? max_dppclk_mhz : - dcn3_1_soc.clock_limits[closest_clk_lvl].dppclk_mhz; - - clock_limits[i].dram_bw_per_chan_gbps = dcn3_1_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps; - clock_limits[i].dscclk_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].dscclk_mhz; - clock_limits[i].dtbclk_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].dtbclk_mhz; - clock_limits[i].phyclk_d18_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].phyclk_d18_mhz; - clock_limits[i].phyclk_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].phyclk_mhz; - } - for (i = 0; i < clk_table->num_entries; i++) - dcn3_1_soc.clock_limits[i] = clock_limits[i]; - if (clk_table->num_entries) { - dcn3_1_soc.num_states = clk_table->num_entries; - } - } - - dcn3_1_soc.dispclk_dppclk_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz / 1000.0; - dc->dml.soc.dispclk_dppclk_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz / 1000.0; - - if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) - dml_init_instance(&dc->dml, &dcn3_1_soc, &dcn3_1_ip, DML_PROJECT_DCN31); - else - dml_init_instance(&dc->dml, &dcn3_1_soc, &dcn3_1_ip, DML_PROJECT_DCN31_FPGA); -} - static struct resource_funcs dcn31_res_pool_funcs = { .destroy = dcn31_destroy_resource_pool, .link_enc_create = dcn31_link_encoder_create, diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.h b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.h index 4b7ab21ea15b..1ce6509c1ed1 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.h +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.h @@ -31,6 +31,9 @@ #define TO_DCN31_RES_POOL(pool)\ container_of(pool, struct dcn31_resource_pool, base) +extern struct _vcs_dpi_ip_params_st dcn3_1_ip; +extern struct _vcs_dpi_soc_bounding_box_st dcn3_1_soc; + struct dcn31_resource_pool { struct resource_pool base; }; @@ -47,7 +50,6 @@ int dcn31_populate_dml_pipes_from_context( struct dc *dc, struct dc_state *context, display_e2e_pipe_params_st *pipes, bool fast_validate); -void dcn31_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params); void dcn31_update_soc_for_wm_a(struct dc *dc, struct dc_state *context); struct resource_pool *dcn31_create_resource_pool( diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile index 28978ce62f87..ee911452c048 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile @@ -71,6 +71,7 @@ CFLAGS_$(AMDDALPATH)/dc/dml/dcn30/display_mode_vba_30.o := $(dml_ccflags) $(fram CFLAGS_$(AMDDALPATH)/dc/dml/dcn30/display_rq_dlg_calc_30.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/display_mode_vba_31.o := $(dml_ccflags) $(frame_warn_flag) CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/display_rq_dlg_calc_31.o := $(dml_ccflags) +CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/dcn31_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn301/dcn301_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn302/dcn302_fpu.o := $(dml_ccflags) CFLAGS_$(AMDDALPATH)/dc/dml/dcn303/dcn303_fpu.o := $(dml_ccflags) @@ -114,6 +115,7 @@ DML += dcn20/display_rq_dlg_calc_20v2.o dcn20/display_mode_vba_20v2.o DML += dcn21/display_rq_dlg_calc_21.o dcn21/display_mode_vba_21.o DML += dcn30/display_mode_vba_30.o dcn30/display_rq_dlg_calc_30.o DML += dcn31/display_mode_vba_31.o dcn31/display_rq_dlg_calc_31.o +DML += dcn31/dcn31_fpu.o DML += dcn301/dcn301_fpu.o DML += dcn302/dcn302_fpu.o DML += dcn303/dcn303_fpu.o diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c new file mode 100644 index 000000000000..7ff8fe9e8712 --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c @@ -0,0 +1,406 @@ +/* + * Copyright 2019-2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#include "resource.h" +#include "clk_mgr.h" + +#include "dml/dcn20/dcn20_fpu.h" +#include "dcn31_fpu.h" + +/** + * DOC: DCN31x FPU manipulation Overview + * + * The DCN architecture relies on FPU operations, which require special + * compilation flags and the use of kernel_fpu_begin/end functions; ideally, we + * want to avoid spreading FPU access across multiple files. With this idea in + * mind, this file aims to centralize all DCN3.1.x functions that require FPU + * access in a single place. Code in this file follows the following code + * pattern: + * + * 1. Functions that use FPU operations should be isolated in static functions. + * 2. The FPU functions should have the noinline attribute to ensure anything + * that deals with FP register is contained within this call. + * 3. All function that needs to be accessed outside this file requires a + * public interface that not uses any FPU reference. + * 4. Developers **must not** use DC_FP_START/END in this file, but they need + * to ensure that the caller invokes it before access any function available + * in this file. For this reason, public functions in this file must invoke + * dc_assert_fp_enabled(); + */ + +struct _vcs_dpi_ip_params_st dcn3_1_ip = { + .gpuvm_enable = 1, + .gpuvm_max_page_table_levels = 1, + .hostvm_enable = 1, + .hostvm_max_page_table_levels = 2, + .rob_buffer_size_kbytes = 64, + .det_buffer_size_kbytes = DCN3_1_DEFAULT_DET_SIZE, + .config_return_buffer_size_in_kbytes = 1792, + .compressed_buffer_segment_size_in_kbytes = 64, + .meta_fifo_size_in_kentries = 32, + .zero_size_buffer_entries = 512, + .compbuf_reserved_space_64b = 256, + .compbuf_reserved_space_zs = 64, + .dpp_output_buffer_pixels = 2560, + .opp_output_buffer_lines = 1, + .pixel_chunk_size_kbytes = 8, + .meta_chunk_size_kbytes = 2, + .min_meta_chunk_size_bytes = 256, + .writeback_chunk_size_kbytes = 8, + .ptoi_supported = false, + .num_dsc = 3, + .maximum_dsc_bits_per_component = 10, + .dsc422_native_support = false, + .is_line_buffer_bpp_fixed = true, + .line_buffer_fixed_bpp = 48, + .line_buffer_size_bits = 789504, + .max_line_buffer_lines = 12, + .writeback_interface_buffer_size_kbytes = 90, + .max_num_dpp = 4, + .max_num_otg = 4, + .max_num_hdmi_frl_outputs = 1, + .max_num_wb = 1, + .max_dchub_pscl_bw_pix_per_clk = 4, + .max_pscl_lb_bw_pix_per_clk = 2, + .max_lb_vscl_bw_pix_per_clk = 4, + .max_vscl_hscl_bw_pix_per_clk = 4, + .max_hscl_ratio = 6, + .max_vscl_ratio = 6, + .max_hscl_taps = 8, + .max_vscl_taps = 8, + .dpte_buffer_size_in_pte_reqs_luma = 64, + .dpte_buffer_size_in_pte_reqs_chroma = 34, + .dispclk_ramp_margin_percent = 1, + .max_inter_dcn_tile_repeaters = 8, + .cursor_buffer_size = 16, + .cursor_chunk_size = 2, + .writeback_line_buffer_buffer_size = 0, + .writeback_min_hscl_ratio = 1, + .writeback_min_vscl_ratio = 1, + .writeback_max_hscl_ratio = 1, + .writeback_max_vscl_ratio = 1, + .writeback_max_hscl_taps = 1, + .writeback_max_vscl_taps = 1, + .dppclk_delay_subtotal = 46, + .dppclk_delay_scl = 50, + .dppclk_delay_scl_lb_only = 16, + .dppclk_delay_cnvc_formatter = 27, + .dppclk_delay_cnvc_cursor = 6, + .dispclk_delay_subtotal = 119, + .dynamic_metadata_vm_enabled = false, + .odm_combine_4to1_supported = false, + .dcc_supported = true, +}; + +struct _vcs_dpi_soc_bounding_box_st dcn3_1_soc = { + /*TODO: correct dispclk/dppclk voltage level determination*/ + .clock_limits = { + { + .state = 0, + .dispclk_mhz = 1200.0, + .dppclk_mhz = 1200.0, + .phyclk_mhz = 600.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 186.0, + .dtbclk_mhz = 625.0, + }, + { + .state = 1, + .dispclk_mhz = 1200.0, + .dppclk_mhz = 1200.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 209.0, + .dtbclk_mhz = 625.0, + }, + { + .state = 2, + .dispclk_mhz = 1200.0, + .dppclk_mhz = 1200.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 209.0, + .dtbclk_mhz = 625.0, + }, + { + .state = 3, + .dispclk_mhz = 1200.0, + .dppclk_mhz = 1200.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 371.0, + .dtbclk_mhz = 625.0, + }, + { + .state = 4, + .dispclk_mhz = 1200.0, + .dppclk_mhz = 1200.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 417.0, + .dtbclk_mhz = 625.0, + }, + }, + .num_states = 5, + .sr_exit_time_us = 9.0, + .sr_enter_plus_exit_time_us = 11.0, + .sr_exit_z8_time_us = 442.0, + .sr_enter_plus_exit_z8_time_us = 560.0, + .writeback_latency_us = 12.0, + .dram_channel_width_bytes = 4, + .round_trip_ping_latency_dcfclk_cycles = 106, + .urgent_latency_pixel_data_only_us = 4.0, + .urgent_latency_pixel_mixed_with_vm_data_us = 4.0, + .urgent_latency_vm_data_only_us = 4.0, + .urgent_out_of_order_return_per_channel_pixel_only_bytes = 4096, + .urgent_out_of_order_return_per_channel_pixel_and_vm_bytes = 4096, + .urgent_out_of_order_return_per_channel_vm_only_bytes = 4096, + .pct_ideal_sdp_bw_after_urgent = 80.0, + .pct_ideal_dram_sdp_bw_after_urgent_pixel_only = 65.0, + .pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm = 60.0, + .pct_ideal_dram_sdp_bw_after_urgent_vm_only = 30.0, + .max_avg_sdp_bw_use_normal_percent = 60.0, + .max_avg_dram_bw_use_normal_percent = 60.0, + .fabric_datapath_to_dcn_data_return_bytes = 32, + .return_bus_width_bytes = 64, + .downspread_percent = 0.38, + .dcn_downspread_percent = 0.5, + .gpuvm_min_page_size_bytes = 4096, + .hostvm_min_page_size_bytes = 4096, + .do_urgent_latency_adjustment = false, + .urgent_latency_adjustment_fabric_clock_component_us = 0, + .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, +}; + + +void dcn31_calculate_wm_and_dlg_fp( + struct dc *dc, struct dc_state *context, + display_e2e_pipe_params_st *pipes, + int pipe_cnt, + int vlevel) +{ + int i, pipe_idx; + double dcfclk = context->bw_ctx.dml.vba.DCFCLKState[vlevel][context->bw_ctx.dml.vba.maxMpcComb]; + + dc_assert_fp_enabled(); + + if (context->bw_ctx.dml.soc.min_dcfclk > dcfclk) + dcfclk = context->bw_ctx.dml.soc.min_dcfclk; + + /* We don't recalculate clocks for 0 pipe configs, which can block + * S0i3 as high clocks will block low power states + * Override any clocks that can block S0i3 to min here + */ + if (pipe_cnt == 0) { + context->bw_ctx.bw.dcn.clk.dcfclk_khz = dcfclk; // always should be vlevel 0 + return; + } + + pipes[0].clks_cfg.voltage = vlevel; + pipes[0].clks_cfg.dcfclk_mhz = dcfclk; + pipes[0].clks_cfg.socclk_mhz = context->bw_ctx.dml.soc.clock_limits[vlevel].socclk_mhz; + +#if 0 // TODO + /* Set B: + * TODO + */ + if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B].valid) { + if (vlevel == 0) { + pipes[0].clks_cfg.voltage = 1; + pipes[0].clks_cfg.dcfclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dcfclk_mhz; + } + context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.pstate_latency_us; + context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.sr_enter_plus_exit_time_us; + context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.sr_exit_time_us; + } + context->bw_ctx.bw.dcn.watermarks.b.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_z8_ns = get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.cstate_exit_z8_ns = get_wm_z8_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.b.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + + pipes[0].clks_cfg.voltage = vlevel; + pipes[0].clks_cfg.dcfclk_mhz = dcfclk; + + /* Set C: + * TODO + */ + if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].valid) { + context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.pstate_latency_us; + context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.sr_enter_plus_exit_time_us; + context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.sr_exit_time_us; + } + context->bw_ctx.bw.dcn.watermarks.c.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_z8_ns = get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.cstate_exit_z8_ns = get_wm_z8_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.c.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + + /* Set D: + * TODO + */ + if (dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].valid) { + context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.pstate_latency_us; + context->bw_ctx.dml.soc.sr_enter_plus_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.sr_enter_plus_exit_time_us; + context->bw_ctx.dml.soc.sr_exit_time_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.sr_exit_time_us; + } + context->bw_ctx.bw.dcn.watermarks.d.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_z8_ns = get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.cstate_pstate.cstate_exit_z8_ns = get_wm_z8_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.d.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; +#endif + + /* Set A: + * All clocks min required + * + * Set A calculated last so that following calculations are based on Set A + */ + dc->res_pool->funcs->update_soc_for_wm_a(dc, context); + context->bw_ctx.bw.dcn.watermarks.a.urgent_ns = get_wm_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns = get_wm_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns = get_wm_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns = get_wm_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_z8_ns = get_wm_z8_stutter_enter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.cstate_pstate.cstate_exit_z8_ns = get_wm_z8_stutter_exit(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.pte_meta_urgent_ns = get_wm_memory_trip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_nom = get_fraction_of_urgent_bandwidth(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.frac_urg_bw_flip = get_fraction_of_urgent_bandwidth_imm_flip(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + context->bw_ctx.bw.dcn.watermarks.a.urgent_latency_ns = get_urgent_latency(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; + /* TODO: remove: */ + context->bw_ctx.bw.dcn.watermarks.b = context->bw_ctx.bw.dcn.watermarks.a; + context->bw_ctx.bw.dcn.watermarks.c = context->bw_ctx.bw.dcn.watermarks.a; + context->bw_ctx.bw.dcn.watermarks.d = context->bw_ctx.bw.dcn.watermarks.a; + /* end remove*/ + + for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) { + if (!context->res_ctx.pipe_ctx[i].stream) + continue; + + pipes[pipe_idx].clks_cfg.dispclk_mhz = get_dispclk_calculated(&context->bw_ctx.dml, pipes, pipe_cnt); + pipes[pipe_idx].clks_cfg.dppclk_mhz = get_dppclk_calculated(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx); + + if (dc->config.forced_clocks || dc->debug.max_disp_clk) { + pipes[pipe_idx].clks_cfg.dispclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dispclk_mhz; + pipes[pipe_idx].clks_cfg.dppclk_mhz = context->bw_ctx.dml.soc.clock_limits[0].dppclk_mhz; + } + if (dc->debug.min_disp_clk_khz > pipes[pipe_idx].clks_cfg.dispclk_mhz * 1000) + pipes[pipe_idx].clks_cfg.dispclk_mhz = dc->debug.min_disp_clk_khz / 1000.0; + if (dc->debug.min_dpp_clk_khz > pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000) + pipes[pipe_idx].clks_cfg.dppclk_mhz = dc->debug.min_dpp_clk_khz / 1000.0; + + pipe_idx++; + } + + dcn20_calculate_dlg_params(dc, context, pipes, pipe_cnt, vlevel); +} + +void dcn31_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params) +{ + struct clk_limit_table *clk_table = &bw_params->clk_table; + struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES]; + unsigned int i, closest_clk_lvl; + int j; + + dc_assert_fp_enabled(); + + // Default clock levels are used for diags, which may lead to overclocking. + if (!IS_DIAG_DC(dc->ctx->dce_environment)) { + int max_dispclk_mhz = 0, max_dppclk_mhz = 0; + + dcn3_1_ip.max_num_otg = dc->res_pool->res_cap->num_timing_generator; + dcn3_1_ip.max_num_dpp = dc->res_pool->pipe_count; + dcn3_1_soc.num_chans = bw_params->num_channels; + + ASSERT(clk_table->num_entries); + + /* Prepass to find max clocks independent of voltage level. */ + for (i = 0; i < clk_table->num_entries; ++i) { + if (clk_table->entries[i].dispclk_mhz > max_dispclk_mhz) + max_dispclk_mhz = clk_table->entries[i].dispclk_mhz; + if (clk_table->entries[i].dppclk_mhz > max_dppclk_mhz) + max_dppclk_mhz = clk_table->entries[i].dppclk_mhz; + } + + for (i = 0; i < clk_table->num_entries; i++) { + /* loop backwards*/ + for (closest_clk_lvl = 0, j = dcn3_1_soc.num_states - 1; j >= 0; j--) { + if ((unsigned int) dcn3_1_soc.clock_limits[j].dcfclk_mhz <= clk_table->entries[i].dcfclk_mhz) { + closest_clk_lvl = j; + break; + } + } + + clock_limits[i].state = i; + + /* Clocks dependent on voltage level. */ + clock_limits[i].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz; + clock_limits[i].fabricclk_mhz = clk_table->entries[i].fclk_mhz; + clock_limits[i].socclk_mhz = clk_table->entries[i].socclk_mhz; + clock_limits[i].dram_speed_mts = clk_table->entries[i].memclk_mhz * 2 * clk_table->entries[i].wck_ratio; + + /* Clocks independent of voltage level. */ + clock_limits[i].dispclk_mhz = max_dispclk_mhz ? max_dispclk_mhz : + dcn3_1_soc.clock_limits[closest_clk_lvl].dispclk_mhz; + + clock_limits[i].dppclk_mhz = max_dppclk_mhz ? max_dppclk_mhz : + dcn3_1_soc.clock_limits[closest_clk_lvl].dppclk_mhz; + + clock_limits[i].dram_bw_per_chan_gbps = dcn3_1_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps; + clock_limits[i].dscclk_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].dscclk_mhz; + clock_limits[i].dtbclk_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].dtbclk_mhz; + clock_limits[i].phyclk_d18_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].phyclk_d18_mhz; + clock_limits[i].phyclk_mhz = dcn3_1_soc.clock_limits[closest_clk_lvl].phyclk_mhz; + } + for (i = 0; i < clk_table->num_entries; i++) + dcn3_1_soc.clock_limits[i] = clock_limits[i]; + if (clk_table->num_entries) { + dcn3_1_soc.num_states = clk_table->num_entries; + } + } + + dcn3_1_soc.dispclk_dppclk_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz / 1000.0; + dc->dml.soc.dispclk_dppclk_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz / 1000.0; + + if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) + dml_init_instance(&dc->dml, &dcn3_1_soc, &dcn3_1_ip, DML_PROJECT_DCN31); + else + dml_init_instance(&dc->dml, &dcn3_1_soc, &dcn3_1_ip, DML_PROJECT_DCN31_FPGA); +} diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h new file mode 100644 index 000000000000..baadb5150e7d --- /dev/null +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h @@ -0,0 +1,39 @@ +/* + * Copyright 2019-2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + +#ifndef __DCN31_FPU_H__ +#define __DCN31_FPU_H__ + +#define DCN3_1_DEFAULT_DET_SIZE 384 + +void dcn31_calculate_wm_and_dlg_fp( + struct dc *dc, struct dc_state *context, + display_e2e_pipe_params_st *pipes, + int pipe_cnt, + int vlevel); + +void dcn31_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params); + +#endif /* __DCN31_FPU_H__*/ -- cgit v1.2.3 From fa896297b31b1100ad13f83a7dd83a715e815042 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Mon, 7 Mar 2022 14:48:00 -0100 Subject: drm/amd/display: move FPU related code from dcn315 to dml/dcn31 folder Moves related structs and dcn315_update_bw_bounding_box from dcn315 driver code to dml/dcn31_fpu that centralizes FPU code for DCN 3.1x. Signed-off-by: Melissa Wen Reviewed-by: Alex Hung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn315/Makefile | 26 --- .../drm/amd/display/dc/dcn315/dcn315_resource.c | 232 +-------------------- .../drm/amd/display/dc/dcn315/dcn315_resource.h | 3 + .../gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c | 228 ++++++++++++++++++++ .../gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h | 3 + 5 files changed, 235 insertions(+), 257 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn315/Makefile b/drivers/gpu/drm/amd/display/dc/dcn315/Makefile index c831ad46e81c..59381d24800b 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn315/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dcn315/Makefile @@ -25,32 +25,6 @@ DCN315 = dcn315_resource.o -ifdef CONFIG_X86 -CFLAGS_$(AMDDALPATH)/dc/dcn315/dcn315_resource.o := -msse -endif - -ifdef CONFIG_PPC64 -CFLAGS_$(AMDDALPATH)/dc/dcn315/dcn315_resource.o := -mhard-float -maltivec -endif - -ifdef CONFIG_CC_IS_GCC -ifeq ($(call cc-ifversion, -lt, 0701, y), y) -IS_OLD_GCC = 1 -endif -CFLAGS_$(AMDDALPATH)/dc/dcn315/dcn315_resource.o += -mhard-float -endif - -ifdef CONFIG_X86 -ifdef IS_OLD_GCC -# Stack alignment mismatch, proceed with caution. -# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3 -# (8B stack alignment). -CFLAGS_$(AMDDALPATH)/dc/dcn315/dcn315_resource.o += -mpreferred-stack-boundary=4 -else -CFLAGS_$(AMDDALPATH)/dc/dcn315/dcn315_resource.o += -msse2 -endif -endif - AMD_DAL_DCN315 = $(addprefix $(AMDDALPATH)/dc/dcn315/,$(DCN315)) AMD_DISPLAY_FILES += $(AMD_DAL_DCN315) diff --git a/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c b/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c index 06adb77c206b..fadb89326999 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c @@ -66,6 +66,7 @@ #include "virtual/virtual_stream_encoder.h" #include "dce110/dce110_resource.h" #include "dml/display_mode_vba.h" +#include "dml/dcn31/dcn31_fpu.h" #include "dcn31/dcn31_dccg.h" #include "dcn10/dcn10_resource.h" #include "dcn31/dcn31_panel_cntl.h" @@ -133,158 +134,9 @@ #include "link_enc_cfg.h" -#define DC_LOGGER_INIT(logger) - -#define DCN3_15_DEFAULT_DET_SIZE 192 #define DCN3_15_MAX_DET_SIZE 384 -#define DCN3_15_MIN_COMPBUF_SIZE_KB 128 #define DCN3_15_CRB_SEGMENT_SIZE_KB 64 -struct _vcs_dpi_ip_params_st dcn3_15_ip = { - .gpuvm_enable = 1, - .gpuvm_max_page_table_levels = 1, - .hostvm_enable = 1, - .hostvm_max_page_table_levels = 2, - .rob_buffer_size_kbytes = 64, - .det_buffer_size_kbytes = DCN3_15_DEFAULT_DET_SIZE, - .min_comp_buffer_size_kbytes = DCN3_15_MIN_COMPBUF_SIZE_KB, - .config_return_buffer_size_in_kbytes = 1024, - .compressed_buffer_segment_size_in_kbytes = 64, - .meta_fifo_size_in_kentries = 32, - .zero_size_buffer_entries = 512, - .compbuf_reserved_space_64b = 256, - .compbuf_reserved_space_zs = 64, - .dpp_output_buffer_pixels = 2560, - .opp_output_buffer_lines = 1, - .pixel_chunk_size_kbytes = 8, - .meta_chunk_size_kbytes = 2, - .min_meta_chunk_size_bytes = 256, - .writeback_chunk_size_kbytes = 8, - .ptoi_supported = false, - .num_dsc = 3, - .maximum_dsc_bits_per_component = 10, - .dsc422_native_support = false, - .is_line_buffer_bpp_fixed = true, - .line_buffer_fixed_bpp = 49, - .line_buffer_size_bits = 789504, - .max_line_buffer_lines = 12, - .writeback_interface_buffer_size_kbytes = 90, - .max_num_dpp = 4, - .max_num_otg = 4, - .max_num_hdmi_frl_outputs = 1, - .max_num_wb = 1, - .max_dchub_pscl_bw_pix_per_clk = 4, - .max_pscl_lb_bw_pix_per_clk = 2, - .max_lb_vscl_bw_pix_per_clk = 4, - .max_vscl_hscl_bw_pix_per_clk = 4, - .max_hscl_ratio = 6, - .max_vscl_ratio = 6, - .max_hscl_taps = 8, - .max_vscl_taps = 8, - .dpte_buffer_size_in_pte_reqs_luma = 64, - .dpte_buffer_size_in_pte_reqs_chroma = 34, - .dispclk_ramp_margin_percent = 1, - .max_inter_dcn_tile_repeaters = 9, - .cursor_buffer_size = 16, - .cursor_chunk_size = 2, - .writeback_line_buffer_buffer_size = 0, - .writeback_min_hscl_ratio = 1, - .writeback_min_vscl_ratio = 1, - .writeback_max_hscl_ratio = 1, - .writeback_max_vscl_ratio = 1, - .writeback_max_hscl_taps = 1, - .writeback_max_vscl_taps = 1, - .dppclk_delay_subtotal = 46, - .dppclk_delay_scl = 50, - .dppclk_delay_scl_lb_only = 16, - .dppclk_delay_cnvc_formatter = 27, - .dppclk_delay_cnvc_cursor = 6, - .dispclk_delay_subtotal = 119, - .dynamic_metadata_vm_enabled = false, - .odm_combine_4to1_supported = false, - .dcc_supported = true, -}; - -struct _vcs_dpi_soc_bounding_box_st dcn3_15_soc = { - /*TODO: correct dispclk/dppclk voltage level determination*/ - .clock_limits = { - { - .state = 0, - .dispclk_mhz = 1372.0, - .dppclk_mhz = 1372.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 417.0, - .dtbclk_mhz = 600.0, - }, - { - .state = 1, - .dispclk_mhz = 1372.0, - .dppclk_mhz = 1372.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 417.0, - .dtbclk_mhz = 600.0, - }, - { - .state = 2, - .dispclk_mhz = 1372.0, - .dppclk_mhz = 1372.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 417.0, - .dtbclk_mhz = 600.0, - }, - { - .state = 3, - .dispclk_mhz = 1372.0, - .dppclk_mhz = 1372.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 417.0, - .dtbclk_mhz = 600.0, - }, - { - .state = 4, - .dispclk_mhz = 1372.0, - .dppclk_mhz = 1372.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 417.0, - .dtbclk_mhz = 600.0, - }, - }, - .num_states = 5, - .sr_exit_time_us = 9.0, - .sr_enter_plus_exit_time_us = 11.0, - .sr_exit_z8_time_us = 50.0, - .sr_enter_plus_exit_z8_time_us = 50.0, - .writeback_latency_us = 12.0, - .dram_channel_width_bytes = 4, - .round_trip_ping_latency_dcfclk_cycles = 106, - .urgent_latency_pixel_data_only_us = 4.0, - .urgent_latency_pixel_mixed_with_vm_data_us = 4.0, - .urgent_latency_vm_data_only_us = 4.0, - .urgent_out_of_order_return_per_channel_pixel_only_bytes = 4096, - .urgent_out_of_order_return_per_channel_pixel_and_vm_bytes = 4096, - .urgent_out_of_order_return_per_channel_vm_only_bytes = 4096, - .pct_ideal_sdp_bw_after_urgent = 80.0, - .pct_ideal_dram_sdp_bw_after_urgent_pixel_only = 65.0, - .pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm = 60.0, - .pct_ideal_dram_sdp_bw_after_urgent_vm_only = 30.0, - .max_avg_sdp_bw_use_normal_percent = 60.0, - .max_avg_dram_bw_use_normal_percent = 60.0, - .fabric_datapath_to_dcn_data_return_bytes = 32, - .return_bus_width_bytes = 64, - .downspread_percent = 0.38, - .dcn_downspread_percent = 0.38, - .gpuvm_min_page_size_bytes = 4096, - .hostvm_min_page_size_bytes = 4096, - .do_urgent_latency_adjustment = false, - .urgent_latency_adjustment_fabric_clock_component_us = 0, - .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, -}; - enum dcn31_clk_src_array_id { DCN31_CLK_SRC_PLL0, DCN31_CLK_SRC_PLL1, @@ -1859,88 +1711,6 @@ static struct dc_cap_funcs cap_funcs = { .get_dcc_compression_cap = dcn20_get_dcc_compression_cap }; -static void dcn315_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params) -{ - struct clk_limit_table *clk_table = &bw_params->clk_table; - struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES]; - unsigned int i, closest_clk_lvl; - int max_dispclk_mhz = 0, max_dppclk_mhz = 0; - int j; - - // Default clock levels are used for diags, which may lead to overclocking. - if (!IS_DIAG_DC(dc->ctx->dce_environment)) { - - dcn3_15_ip.max_num_otg = dc->res_pool->res_cap->num_timing_generator; - dcn3_15_ip.max_num_dpp = dc->res_pool->pipe_count; - dcn3_15_soc.num_chans = bw_params->num_channels; - - ASSERT(clk_table->num_entries); - - /* Prepass to find max clocks independent of voltage level. */ - for (i = 0; i < clk_table->num_entries; ++i) { - if (clk_table->entries[i].dispclk_mhz > max_dispclk_mhz) - max_dispclk_mhz = clk_table->entries[i].dispclk_mhz; - if (clk_table->entries[i].dppclk_mhz > max_dppclk_mhz) - max_dppclk_mhz = clk_table->entries[i].dppclk_mhz; - } - - for (i = 0; i < clk_table->num_entries; i++) { - /* loop backwards*/ - for (closest_clk_lvl = 0, j = dcn3_15_soc.num_states - 1; j >= 0; j--) { - if ((unsigned int) dcn3_15_soc.clock_limits[j].dcfclk_mhz <= clk_table->entries[i].dcfclk_mhz) { - closest_clk_lvl = j; - break; - } - } - if (clk_table->num_entries == 1) { - /*smu gives one DPM level, let's take the highest one*/ - closest_clk_lvl = dcn3_15_soc.num_states - 1; - } - - clock_limits[i].state = i; - - /* Clocks dependent on voltage level. */ - clock_limits[i].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz; - if (clk_table->num_entries == 1 && - clock_limits[i].dcfclk_mhz < dcn3_15_soc.clock_limits[closest_clk_lvl].dcfclk_mhz) { - /*SMU fix not released yet*/ - clock_limits[i].dcfclk_mhz = dcn3_15_soc.clock_limits[closest_clk_lvl].dcfclk_mhz; - } - clock_limits[i].fabricclk_mhz = clk_table->entries[i].fclk_mhz; - clock_limits[i].socclk_mhz = clk_table->entries[i].socclk_mhz; - clock_limits[i].dram_speed_mts = clk_table->entries[i].memclk_mhz * 2 * clk_table->entries[i].wck_ratio; - - /* Clocks independent of voltage level. */ - clock_limits[i].dispclk_mhz = max_dispclk_mhz ? max_dispclk_mhz : - dcn3_15_soc.clock_limits[closest_clk_lvl].dispclk_mhz; - - clock_limits[i].dppclk_mhz = max_dppclk_mhz ? max_dppclk_mhz : - dcn3_15_soc.clock_limits[closest_clk_lvl].dppclk_mhz; - - clock_limits[i].dram_bw_per_chan_gbps = dcn3_15_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps; - clock_limits[i].dscclk_mhz = dcn3_15_soc.clock_limits[closest_clk_lvl].dscclk_mhz; - clock_limits[i].dtbclk_mhz = dcn3_15_soc.clock_limits[closest_clk_lvl].dtbclk_mhz; - clock_limits[i].phyclk_d18_mhz = dcn3_15_soc.clock_limits[closest_clk_lvl].phyclk_d18_mhz; - clock_limits[i].phyclk_mhz = dcn3_15_soc.clock_limits[closest_clk_lvl].phyclk_mhz; - } - for (i = 0; i < clk_table->num_entries; i++) - dcn3_15_soc.clock_limits[i] = clock_limits[i]; - if (clk_table->num_entries) { - dcn3_15_soc.num_states = clk_table->num_entries; - } - } - - if (max_dispclk_mhz) { - dcn3_15_soc.dispclk_dppclk_vco_speed_mhz = max_dispclk_mhz * 2; - dc->dml.soc.dispclk_dppclk_vco_speed_mhz = max_dispclk_mhz * 2; - } - - if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) - dml_init_instance(&dc->dml, &dcn3_15_soc, &dcn3_15_ip, DML_PROJECT_DCN31); - else - dml_init_instance(&dc->dml, &dcn3_15_soc, &dcn3_15_ip, DML_PROJECT_DCN31_FPGA); -} - static struct resource_funcs dcn315_res_pool_funcs = { .destroy = dcn315_destroy_resource_pool, .link_enc_create = dcn31_link_encoder_create, diff --git a/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.h b/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.h index f3a36820a31f..39929fa67a51 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.h +++ b/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.h @@ -31,6 +31,9 @@ #define TO_DCN315_RES_POOL(pool)\ container_of(pool, struct dcn315_resource_pool, base) +extern struct _vcs_dpi_ip_params_st dcn3_15_ip; +extern struct _vcs_dpi_ip_params_st dcn3_15_soc; + struct dcn315_resource_pool { struct resource_pool base; }; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c index 7ff8fe9e8712..f70b47ef850c 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c @@ -194,6 +194,150 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_1_soc = { .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, }; +struct _vcs_dpi_ip_params_st dcn3_15_ip = { + .gpuvm_enable = 1, + .gpuvm_max_page_table_levels = 1, + .hostvm_enable = 1, + .hostvm_max_page_table_levels = 2, + .rob_buffer_size_kbytes = 64, + .det_buffer_size_kbytes = DCN3_15_DEFAULT_DET_SIZE, + .min_comp_buffer_size_kbytes = DCN3_15_MIN_COMPBUF_SIZE_KB, + .config_return_buffer_size_in_kbytes = 1024, + .compressed_buffer_segment_size_in_kbytes = 64, + .meta_fifo_size_in_kentries = 32, + .zero_size_buffer_entries = 512, + .compbuf_reserved_space_64b = 256, + .compbuf_reserved_space_zs = 64, + .dpp_output_buffer_pixels = 2560, + .opp_output_buffer_lines = 1, + .pixel_chunk_size_kbytes = 8, + .meta_chunk_size_kbytes = 2, + .min_meta_chunk_size_bytes = 256, + .writeback_chunk_size_kbytes = 8, + .ptoi_supported = false, + .num_dsc = 3, + .maximum_dsc_bits_per_component = 10, + .dsc422_native_support = false, + .is_line_buffer_bpp_fixed = true, + .line_buffer_fixed_bpp = 49, + .line_buffer_size_bits = 789504, + .max_line_buffer_lines = 12, + .writeback_interface_buffer_size_kbytes = 90, + .max_num_dpp = 4, + .max_num_otg = 4, + .max_num_hdmi_frl_outputs = 1, + .max_num_wb = 1, + .max_dchub_pscl_bw_pix_per_clk = 4, + .max_pscl_lb_bw_pix_per_clk = 2, + .max_lb_vscl_bw_pix_per_clk = 4, + .max_vscl_hscl_bw_pix_per_clk = 4, + .max_hscl_ratio = 6, + .max_vscl_ratio = 6, + .max_hscl_taps = 8, + .max_vscl_taps = 8, + .dpte_buffer_size_in_pte_reqs_luma = 64, + .dpte_buffer_size_in_pte_reqs_chroma = 34, + .dispclk_ramp_margin_percent = 1, + .max_inter_dcn_tile_repeaters = 9, + .cursor_buffer_size = 16, + .cursor_chunk_size = 2, + .writeback_line_buffer_buffer_size = 0, + .writeback_min_hscl_ratio = 1, + .writeback_min_vscl_ratio = 1, + .writeback_max_hscl_ratio = 1, + .writeback_max_vscl_ratio = 1, + .writeback_max_hscl_taps = 1, + .writeback_max_vscl_taps = 1, + .dppclk_delay_subtotal = 46, + .dppclk_delay_scl = 50, + .dppclk_delay_scl_lb_only = 16, + .dppclk_delay_cnvc_formatter = 27, + .dppclk_delay_cnvc_cursor = 6, + .dispclk_delay_subtotal = 119, + .dynamic_metadata_vm_enabled = false, + .odm_combine_4to1_supported = false, + .dcc_supported = true, +}; + +struct _vcs_dpi_soc_bounding_box_st dcn3_15_soc = { + /*TODO: correct dispclk/dppclk voltage level determination*/ + .clock_limits = { + { + .state = 0, + .dispclk_mhz = 1372.0, + .dppclk_mhz = 1372.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 417.0, + .dtbclk_mhz = 600.0, + }, + { + .state = 1, + .dispclk_mhz = 1372.0, + .dppclk_mhz = 1372.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 417.0, + .dtbclk_mhz = 600.0, + }, + { + .state = 2, + .dispclk_mhz = 1372.0, + .dppclk_mhz = 1372.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 417.0, + .dtbclk_mhz = 600.0, + }, + { + .state = 3, + .dispclk_mhz = 1372.0, + .dppclk_mhz = 1372.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 417.0, + .dtbclk_mhz = 600.0, + }, + { + .state = 4, + .dispclk_mhz = 1372.0, + .dppclk_mhz = 1372.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 417.0, + .dtbclk_mhz = 600.0, + }, + }, + .num_states = 5, + .sr_exit_time_us = 9.0, + .sr_enter_plus_exit_time_us = 11.0, + .sr_exit_z8_time_us = 50.0, + .sr_enter_plus_exit_z8_time_us = 50.0, + .writeback_latency_us = 12.0, + .dram_channel_width_bytes = 4, + .round_trip_ping_latency_dcfclk_cycles = 106, + .urgent_latency_pixel_data_only_us = 4.0, + .urgent_latency_pixel_mixed_with_vm_data_us = 4.0, + .urgent_latency_vm_data_only_us = 4.0, + .urgent_out_of_order_return_per_channel_pixel_only_bytes = 4096, + .urgent_out_of_order_return_per_channel_pixel_and_vm_bytes = 4096, + .urgent_out_of_order_return_per_channel_vm_only_bytes = 4096, + .pct_ideal_sdp_bw_after_urgent = 80.0, + .pct_ideal_dram_sdp_bw_after_urgent_pixel_only = 65.0, + .pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm = 60.0, + .pct_ideal_dram_sdp_bw_after_urgent_vm_only = 30.0, + .max_avg_sdp_bw_use_normal_percent = 60.0, + .max_avg_dram_bw_use_normal_percent = 60.0, + .fabric_datapath_to_dcn_data_return_bytes = 32, + .return_bus_width_bytes = 64, + .downspread_percent = 0.38, + .dcn_downspread_percent = 0.38, + .gpuvm_min_page_size_bytes = 4096, + .hostvm_min_page_size_bytes = 4096, + .do_urgent_latency_adjustment = false, + .urgent_latency_adjustment_fabric_clock_component_us = 0, + .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, +}; void dcn31_calculate_wm_and_dlg_fp( struct dc *dc, struct dc_state *context, @@ -404,3 +548,87 @@ void dcn31_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params else dml_init_instance(&dc->dml, &dcn3_1_soc, &dcn3_1_ip, DML_PROJECT_DCN31_FPGA); } + +void dcn315_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params) +{ + struct clk_limit_table *clk_table = &bw_params->clk_table; + struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES]; + unsigned int i, closest_clk_lvl; + int max_dispclk_mhz = 0, max_dppclk_mhz = 0; + int j; + + dc_assert_fp_enabled(); + + // Default clock levels are used for diags, which may lead to overclocking. + if (!IS_DIAG_DC(dc->ctx->dce_environment)) { + + dcn3_15_ip.max_num_otg = dc->res_pool->res_cap->num_timing_generator; + dcn3_15_ip.max_num_dpp = dc->res_pool->pipe_count; + dcn3_15_soc.num_chans = bw_params->num_channels; + + ASSERT(clk_table->num_entries); + + /* Prepass to find max clocks independent of voltage level. */ + for (i = 0; i < clk_table->num_entries; ++i) { + if (clk_table->entries[i].dispclk_mhz > max_dispclk_mhz) + max_dispclk_mhz = clk_table->entries[i].dispclk_mhz; + if (clk_table->entries[i].dppclk_mhz > max_dppclk_mhz) + max_dppclk_mhz = clk_table->entries[i].dppclk_mhz; + } + + for (i = 0; i < clk_table->num_entries; i++) { + /* loop backwards*/ + for (closest_clk_lvl = 0, j = dcn3_15_soc.num_states - 1; j >= 0; j--) { + if ((unsigned int) dcn3_15_soc.clock_limits[j].dcfclk_mhz <= clk_table->entries[i].dcfclk_mhz) { + closest_clk_lvl = j; + break; + } + } + if (clk_table->num_entries == 1) { + /*smu gives one DPM level, let's take the highest one*/ + closest_clk_lvl = dcn3_15_soc.num_states - 1; + } + + clock_limits[i].state = i; + + /* Clocks dependent on voltage level. */ + clock_limits[i].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz; + if (clk_table->num_entries == 1 && + clock_limits[i].dcfclk_mhz < dcn3_15_soc.clock_limits[closest_clk_lvl].dcfclk_mhz) { + /*SMU fix not released yet*/ + clock_limits[i].dcfclk_mhz = dcn3_15_soc.clock_limits[closest_clk_lvl].dcfclk_mhz; + } + clock_limits[i].fabricclk_mhz = clk_table->entries[i].fclk_mhz; + clock_limits[i].socclk_mhz = clk_table->entries[i].socclk_mhz; + clock_limits[i].dram_speed_mts = clk_table->entries[i].memclk_mhz * 2 * clk_table->entries[i].wck_ratio; + + /* Clocks independent of voltage level. */ + clock_limits[i].dispclk_mhz = max_dispclk_mhz ? max_dispclk_mhz : + dcn3_15_soc.clock_limits[closest_clk_lvl].dispclk_mhz; + + clock_limits[i].dppclk_mhz = max_dppclk_mhz ? max_dppclk_mhz : + dcn3_15_soc.clock_limits[closest_clk_lvl].dppclk_mhz; + + clock_limits[i].dram_bw_per_chan_gbps = dcn3_15_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps; + clock_limits[i].dscclk_mhz = dcn3_15_soc.clock_limits[closest_clk_lvl].dscclk_mhz; + clock_limits[i].dtbclk_mhz = dcn3_15_soc.clock_limits[closest_clk_lvl].dtbclk_mhz; + clock_limits[i].phyclk_d18_mhz = dcn3_15_soc.clock_limits[closest_clk_lvl].phyclk_d18_mhz; + clock_limits[i].phyclk_mhz = dcn3_15_soc.clock_limits[closest_clk_lvl].phyclk_mhz; + } + for (i = 0; i < clk_table->num_entries; i++) + dcn3_15_soc.clock_limits[i] = clock_limits[i]; + if (clk_table->num_entries) { + dcn3_15_soc.num_states = clk_table->num_entries; + } + } + + if (max_dispclk_mhz) { + dcn3_15_soc.dispclk_dppclk_vco_speed_mhz = max_dispclk_mhz * 2; + dc->dml.soc.dispclk_dppclk_vco_speed_mhz = max_dispclk_mhz * 2; + } + + if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) + dml_init_instance(&dc->dml, &dcn3_15_soc, &dcn3_15_ip, DML_PROJECT_DCN31); + else + dml_init_instance(&dc->dml, &dcn3_15_soc, &dcn3_15_ip, DML_PROJECT_DCN31_FPGA); +} diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h index baadb5150e7d..b15b587cf8c4 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h @@ -27,6 +27,8 @@ #define __DCN31_FPU_H__ #define DCN3_1_DEFAULT_DET_SIZE 384 +#define DCN3_15_DEFAULT_DET_SIZE 192 +#define DCN3_15_MIN_COMPBUF_SIZE_KB 128 void dcn31_calculate_wm_and_dlg_fp( struct dc *dc, struct dc_state *context, @@ -35,5 +37,6 @@ void dcn31_calculate_wm_and_dlg_fp( int vlevel); void dcn31_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params); +void dcn315_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params); #endif /* __DCN31_FPU_H__*/ -- cgit v1.2.3 From 3f8951cc123f102a4e21d61390993bc7d142b723 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Mon, 7 Mar 2022 14:48:01 -0100 Subject: drm/amd/display: move FPU related code from dcn316 to dml/dcn31 folder Moves FPU-related structs and dcn316_update_bw_bounding_box from dcn316 driver to dml/dcn31 that centralize FPU operations for DCN 3.1x Signed-off-by: Melissa Wen Reviewed-by: Alex Hung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn316/Makefile | 26 --- .../drm/amd/display/dc/dcn316/dcn316_resource.c | 231 +-------------------- .../drm/amd/display/dc/dcn316/dcn316_resource.h | 3 + .../gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c | 229 ++++++++++++++++++++ .../gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h | 2 + 5 files changed, 235 insertions(+), 256 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn316/Makefile b/drivers/gpu/drm/amd/display/dc/dcn316/Makefile index cd87b687c5e2..819d44a9439b 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn316/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dcn316/Makefile @@ -25,32 +25,6 @@ DCN316 = dcn316_resource.o -ifdef CONFIG_X86 -CFLAGS_$(AMDDALPATH)/dc/dcn316/dcn316_resource.o := -msse -endif - -ifdef CONFIG_PPC64 -CFLAGS_$(AMDDALPATH)/dc/dcn316/dcn316_resource.o := -mhard-float -maltivec -endif - -ifdef CONFIG_CC_IS_GCC -ifeq ($(call cc-ifversion, -lt, 0701, y), y) -IS_OLD_GCC = 1 -endif -CFLAGS_$(AMDDALPATH)/dc/dcn316/dcn316_resource.o += -mhard-float -endif - -ifdef CONFIG_X86 -ifdef IS_OLD_GCC -# Stack alignment mismatch, proceed with caution. -# GCC < 7.1 cannot compile code using `double` and -mpreferred-stack-boundary=3 -# (8B stack alignment). -CFLAGS_$(AMDDALPATH)/dc/dcn316/dcn316_resource.o += -mpreferred-stack-boundary=4 -else -CFLAGS_$(AMDDALPATH)/dc/dcn316/dcn316_resource.o += -msse2 -endif -endif - AMD_DAL_DCN316 = $(addprefix $(AMDDALPATH)/dc/dcn316/,$(DCN316)) AMD_DISPLAY_FILES += $(AMD_DAL_DCN316) diff --git a/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c b/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c index 8decc3ccf8ca..d73145dab173 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c @@ -66,6 +66,7 @@ #include "virtual/virtual_stream_encoder.h" #include "dce110/dce110_resource.h" #include "dml/display_mode_vba.h" +#include "dml/dcn31/dcn31_fpu.h" #include "dcn31/dcn31_dccg.h" #include "dcn10/dcn10_resource.h" #include "dcn31/dcn31_panel_cntl.h" @@ -123,157 +124,10 @@ #include "link_enc_cfg.h" -#define DC_LOGGER_INIT(logger) - -#define DCN3_16_DEFAULT_DET_SIZE 192 #define DCN3_16_MAX_DET_SIZE 384 #define DCN3_16_MIN_COMPBUF_SIZE_KB 128 #define DCN3_16_CRB_SEGMENT_SIZE_KB 64 -struct _vcs_dpi_ip_params_st dcn3_16_ip = { - .gpuvm_enable = 1, - .gpuvm_max_page_table_levels = 1, - .hostvm_enable = 1, - .hostvm_max_page_table_levels = 2, - .rob_buffer_size_kbytes = 64, - .det_buffer_size_kbytes = DCN3_16_DEFAULT_DET_SIZE, - .config_return_buffer_size_in_kbytes = 1024, - .compressed_buffer_segment_size_in_kbytes = 64, - .meta_fifo_size_in_kentries = 32, - .zero_size_buffer_entries = 512, - .compbuf_reserved_space_64b = 256, - .compbuf_reserved_space_zs = 64, - .dpp_output_buffer_pixels = 2560, - .opp_output_buffer_lines = 1, - .pixel_chunk_size_kbytes = 8, - .meta_chunk_size_kbytes = 2, - .min_meta_chunk_size_bytes = 256, - .writeback_chunk_size_kbytes = 8, - .ptoi_supported = false, - .num_dsc = 3, - .maximum_dsc_bits_per_component = 10, - .dsc422_native_support = false, - .is_line_buffer_bpp_fixed = true, - .line_buffer_fixed_bpp = 48, - .line_buffer_size_bits = 789504, - .max_line_buffer_lines = 12, - .writeback_interface_buffer_size_kbytes = 90, - .max_num_dpp = 4, - .max_num_otg = 4, - .max_num_hdmi_frl_outputs = 1, - .max_num_wb = 1, - .max_dchub_pscl_bw_pix_per_clk = 4, - .max_pscl_lb_bw_pix_per_clk = 2, - .max_lb_vscl_bw_pix_per_clk = 4, - .max_vscl_hscl_bw_pix_per_clk = 4, - .max_hscl_ratio = 6, - .max_vscl_ratio = 6, - .max_hscl_taps = 8, - .max_vscl_taps = 8, - .dpte_buffer_size_in_pte_reqs_luma = 64, - .dpte_buffer_size_in_pte_reqs_chroma = 34, - .dispclk_ramp_margin_percent = 1, - .max_inter_dcn_tile_repeaters = 8, - .cursor_buffer_size = 16, - .cursor_chunk_size = 2, - .writeback_line_buffer_buffer_size = 0, - .writeback_min_hscl_ratio = 1, - .writeback_min_vscl_ratio = 1, - .writeback_max_hscl_ratio = 1, - .writeback_max_vscl_ratio = 1, - .writeback_max_hscl_taps = 1, - .writeback_max_vscl_taps = 1, - .dppclk_delay_subtotal = 46, - .dppclk_delay_scl = 50, - .dppclk_delay_scl_lb_only = 16, - .dppclk_delay_cnvc_formatter = 27, - .dppclk_delay_cnvc_cursor = 6, - .dispclk_delay_subtotal = 119, - .dynamic_metadata_vm_enabled = false, - .odm_combine_4to1_supported = false, - .dcc_supported = true, -}; - -struct _vcs_dpi_soc_bounding_box_st dcn3_16_soc = { - /*TODO: correct dispclk/dppclk voltage level determination*/ - .clock_limits = { - { - .state = 0, - .dispclk_mhz = 556.0, - .dppclk_mhz = 556.0, - .phyclk_mhz = 600.0, - .phyclk_d18_mhz = 445.0, - .dscclk_mhz = 186.0, - .dtbclk_mhz = 625.0, - }, - { - .state = 1, - .dispclk_mhz = 625.0, - .dppclk_mhz = 625.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 209.0, - .dtbclk_mhz = 625.0, - }, - { - .state = 2, - .dispclk_mhz = 625.0, - .dppclk_mhz = 625.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 209.0, - .dtbclk_mhz = 625.0, - }, - { - .state = 3, - .dispclk_mhz = 1112.0, - .dppclk_mhz = 1112.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 371.0, - .dtbclk_mhz = 625.0, - }, - { - .state = 4, - .dispclk_mhz = 1250.0, - .dppclk_mhz = 1250.0, - .phyclk_mhz = 810.0, - .phyclk_d18_mhz = 667.0, - .dscclk_mhz = 417.0, - .dtbclk_mhz = 625.0, - }, - }, - .num_states = 5, - .sr_exit_time_us = 9.0, - .sr_enter_plus_exit_time_us = 11.0, - .sr_exit_z8_time_us = 442.0, - .sr_enter_plus_exit_z8_time_us = 560.0, - .writeback_latency_us = 12.0, - .dram_channel_width_bytes = 4, - .round_trip_ping_latency_dcfclk_cycles = 106, - .urgent_latency_pixel_data_only_us = 4.0, - .urgent_latency_pixel_mixed_with_vm_data_us = 4.0, - .urgent_latency_vm_data_only_us = 4.0, - .urgent_out_of_order_return_per_channel_pixel_only_bytes = 4096, - .urgent_out_of_order_return_per_channel_pixel_and_vm_bytes = 4096, - .urgent_out_of_order_return_per_channel_vm_only_bytes = 4096, - .pct_ideal_sdp_bw_after_urgent = 80.0, - .pct_ideal_dram_sdp_bw_after_urgent_pixel_only = 65.0, - .pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm = 60.0, - .pct_ideal_dram_sdp_bw_after_urgent_vm_only = 30.0, - .max_avg_sdp_bw_use_normal_percent = 60.0, - .max_avg_dram_bw_use_normal_percent = 60.0, - .fabric_datapath_to_dcn_data_return_bytes = 32, - .return_bus_width_bytes = 64, - .downspread_percent = 0.38, - .dcn_downspread_percent = 0.5, - .gpuvm_min_page_size_bytes = 4096, - .hostvm_min_page_size_bytes = 4096, - .do_urgent_latency_adjustment = false, - .urgent_latency_adjustment_fabric_clock_component_us = 0, - .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, -}; - enum dcn31_clk_src_array_id { DCN31_CLK_SRC_PLL0, DCN31_CLK_SRC_PLL1, @@ -1859,89 +1713,6 @@ static struct dc_cap_funcs cap_funcs = { .get_dcc_compression_cap = dcn20_get_dcc_compression_cap }; -static void dcn316_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params) -{ - struct clk_limit_table *clk_table = &bw_params->clk_table; - struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES]; - unsigned int i, closest_clk_lvl; - int max_dispclk_mhz = 0, max_dppclk_mhz = 0; - int j; - - // Default clock levels are used for diags, which may lead to overclocking. - if (!IS_DIAG_DC(dc->ctx->dce_environment)) { - - dcn3_16_ip.max_num_otg = dc->res_pool->res_cap->num_timing_generator; - dcn3_16_ip.max_num_dpp = dc->res_pool->pipe_count; - dcn3_16_soc.num_chans = bw_params->num_channels; - - ASSERT(clk_table->num_entries); - - /* Prepass to find max clocks independent of voltage level. */ - for (i = 0; i < clk_table->num_entries; ++i) { - if (clk_table->entries[i].dispclk_mhz > max_dispclk_mhz) - max_dispclk_mhz = clk_table->entries[i].dispclk_mhz; - if (clk_table->entries[i].dppclk_mhz > max_dppclk_mhz) - max_dppclk_mhz = clk_table->entries[i].dppclk_mhz; - } - - for (i = 0; i < clk_table->num_entries; i++) { - /* loop backwards*/ - for (closest_clk_lvl = 0, j = dcn3_16_soc.num_states - 1; j >= 0; j--) { - if ((unsigned int) dcn3_16_soc.clock_limits[j].dcfclk_mhz <= clk_table->entries[i].dcfclk_mhz) { - closest_clk_lvl = j; - break; - } - } - // Ported from DCN315 - if (clk_table->num_entries == 1) { - /*smu gives one DPM level, let's take the highest one*/ - closest_clk_lvl = dcn3_16_soc.num_states - 1; - } - - clock_limits[i].state = i; - - /* Clocks dependent on voltage level. */ - clock_limits[i].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz; - if (clk_table->num_entries == 1 && - clock_limits[i].dcfclk_mhz < dcn3_16_soc.clock_limits[closest_clk_lvl].dcfclk_mhz) { - /*SMU fix not released yet*/ - clock_limits[i].dcfclk_mhz = dcn3_16_soc.clock_limits[closest_clk_lvl].dcfclk_mhz; - } - clock_limits[i].fabricclk_mhz = clk_table->entries[i].fclk_mhz; - clock_limits[i].socclk_mhz = clk_table->entries[i].socclk_mhz; - clock_limits[i].dram_speed_mts = clk_table->entries[i].memclk_mhz * 2 * clk_table->entries[i].wck_ratio; - - /* Clocks independent of voltage level. */ - clock_limits[i].dispclk_mhz = max_dispclk_mhz ? max_dispclk_mhz : - dcn3_16_soc.clock_limits[closest_clk_lvl].dispclk_mhz; - - clock_limits[i].dppclk_mhz = max_dppclk_mhz ? max_dppclk_mhz : - dcn3_16_soc.clock_limits[closest_clk_lvl].dppclk_mhz; - - clock_limits[i].dram_bw_per_chan_gbps = dcn3_16_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps; - clock_limits[i].dscclk_mhz = dcn3_16_soc.clock_limits[closest_clk_lvl].dscclk_mhz; - clock_limits[i].dtbclk_mhz = dcn3_16_soc.clock_limits[closest_clk_lvl].dtbclk_mhz; - clock_limits[i].phyclk_d18_mhz = dcn3_16_soc.clock_limits[closest_clk_lvl].phyclk_d18_mhz; - clock_limits[i].phyclk_mhz = dcn3_16_soc.clock_limits[closest_clk_lvl].phyclk_mhz; - } - for (i = 0; i < clk_table->num_entries; i++) - dcn3_16_soc.clock_limits[i] = clock_limits[i]; - if (clk_table->num_entries) { - dcn3_16_soc.num_states = clk_table->num_entries; - } - } - - if (max_dispclk_mhz) { - dcn3_16_soc.dispclk_dppclk_vco_speed_mhz = max_dispclk_mhz * 2; - dc->dml.soc.dispclk_dppclk_vco_speed_mhz = max_dispclk_mhz * 2; - } - - if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) - dml_init_instance(&dc->dml, &dcn3_16_soc, &dcn3_16_ip, DML_PROJECT_DCN31); - else - dml_init_instance(&dc->dml, &dcn3_16_soc, &dcn3_16_ip, DML_PROJECT_DCN31_FPGA); -} - static struct resource_funcs dcn316_res_pool_funcs = { .destroy = dcn316_destroy_resource_pool, .link_enc_create = dcn31_link_encoder_create, diff --git a/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.h b/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.h index 9d0d60cb9482..0dc5a6c13ae7 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.h +++ b/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.h @@ -31,6 +31,9 @@ #define TO_DCN316_RES_POOL(pool)\ container_of(pool, struct dcn316_resource_pool, base) +extern struct _vcs_dpi_ip_params_st dcn3_16_ip; +extern struct _vcs_dpi_ip_params_st dcn3_16_soc; + struct dcn316_resource_pool { struct resource_pool base; }; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c index f70b47ef850c..a0a2e125c9c8 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.c @@ -339,6 +339,150 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_15_soc = { .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, }; +struct _vcs_dpi_ip_params_st dcn3_16_ip = { + .gpuvm_enable = 1, + .gpuvm_max_page_table_levels = 1, + .hostvm_enable = 1, + .hostvm_max_page_table_levels = 2, + .rob_buffer_size_kbytes = 64, + .det_buffer_size_kbytes = DCN3_16_DEFAULT_DET_SIZE, + .config_return_buffer_size_in_kbytes = 1024, + .compressed_buffer_segment_size_in_kbytes = 64, + .meta_fifo_size_in_kentries = 32, + .zero_size_buffer_entries = 512, + .compbuf_reserved_space_64b = 256, + .compbuf_reserved_space_zs = 64, + .dpp_output_buffer_pixels = 2560, + .opp_output_buffer_lines = 1, + .pixel_chunk_size_kbytes = 8, + .meta_chunk_size_kbytes = 2, + .min_meta_chunk_size_bytes = 256, + .writeback_chunk_size_kbytes = 8, + .ptoi_supported = false, + .num_dsc = 3, + .maximum_dsc_bits_per_component = 10, + .dsc422_native_support = false, + .is_line_buffer_bpp_fixed = true, + .line_buffer_fixed_bpp = 48, + .line_buffer_size_bits = 789504, + .max_line_buffer_lines = 12, + .writeback_interface_buffer_size_kbytes = 90, + .max_num_dpp = 4, + .max_num_otg = 4, + .max_num_hdmi_frl_outputs = 1, + .max_num_wb = 1, + .max_dchub_pscl_bw_pix_per_clk = 4, + .max_pscl_lb_bw_pix_per_clk = 2, + .max_lb_vscl_bw_pix_per_clk = 4, + .max_vscl_hscl_bw_pix_per_clk = 4, + .max_hscl_ratio = 6, + .max_vscl_ratio = 6, + .max_hscl_taps = 8, + .max_vscl_taps = 8, + .dpte_buffer_size_in_pte_reqs_luma = 64, + .dpte_buffer_size_in_pte_reqs_chroma = 34, + .dispclk_ramp_margin_percent = 1, + .max_inter_dcn_tile_repeaters = 8, + .cursor_buffer_size = 16, + .cursor_chunk_size = 2, + .writeback_line_buffer_buffer_size = 0, + .writeback_min_hscl_ratio = 1, + .writeback_min_vscl_ratio = 1, + .writeback_max_hscl_ratio = 1, + .writeback_max_vscl_ratio = 1, + .writeback_max_hscl_taps = 1, + .writeback_max_vscl_taps = 1, + .dppclk_delay_subtotal = 46, + .dppclk_delay_scl = 50, + .dppclk_delay_scl_lb_only = 16, + .dppclk_delay_cnvc_formatter = 27, + .dppclk_delay_cnvc_cursor = 6, + .dispclk_delay_subtotal = 119, + .dynamic_metadata_vm_enabled = false, + .odm_combine_4to1_supported = false, + .dcc_supported = true, +}; + +struct _vcs_dpi_soc_bounding_box_st dcn3_16_soc = { + /*TODO: correct dispclk/dppclk voltage level determination*/ + .clock_limits = { + { + .state = 0, + .dispclk_mhz = 556.0, + .dppclk_mhz = 556.0, + .phyclk_mhz = 600.0, + .phyclk_d18_mhz = 445.0, + .dscclk_mhz = 186.0, + .dtbclk_mhz = 625.0, + }, + { + .state = 1, + .dispclk_mhz = 625.0, + .dppclk_mhz = 625.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 209.0, + .dtbclk_mhz = 625.0, + }, + { + .state = 2, + .dispclk_mhz = 625.0, + .dppclk_mhz = 625.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 209.0, + .dtbclk_mhz = 625.0, + }, + { + .state = 3, + .dispclk_mhz = 1112.0, + .dppclk_mhz = 1112.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 371.0, + .dtbclk_mhz = 625.0, + }, + { + .state = 4, + .dispclk_mhz = 1250.0, + .dppclk_mhz = 1250.0, + .phyclk_mhz = 810.0, + .phyclk_d18_mhz = 667.0, + .dscclk_mhz = 417.0, + .dtbclk_mhz = 625.0, + }, + }, + .num_states = 5, + .sr_exit_time_us = 9.0, + .sr_enter_plus_exit_time_us = 11.0, + .sr_exit_z8_time_us = 442.0, + .sr_enter_plus_exit_z8_time_us = 560.0, + .writeback_latency_us = 12.0, + .dram_channel_width_bytes = 4, + .round_trip_ping_latency_dcfclk_cycles = 106, + .urgent_latency_pixel_data_only_us = 4.0, + .urgent_latency_pixel_mixed_with_vm_data_us = 4.0, + .urgent_latency_vm_data_only_us = 4.0, + .urgent_out_of_order_return_per_channel_pixel_only_bytes = 4096, + .urgent_out_of_order_return_per_channel_pixel_and_vm_bytes = 4096, + .urgent_out_of_order_return_per_channel_vm_only_bytes = 4096, + .pct_ideal_sdp_bw_after_urgent = 80.0, + .pct_ideal_dram_sdp_bw_after_urgent_pixel_only = 65.0, + .pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm = 60.0, + .pct_ideal_dram_sdp_bw_after_urgent_vm_only = 30.0, + .max_avg_sdp_bw_use_normal_percent = 60.0, + .max_avg_dram_bw_use_normal_percent = 60.0, + .fabric_datapath_to_dcn_data_return_bytes = 32, + .return_bus_width_bytes = 64, + .downspread_percent = 0.38, + .dcn_downspread_percent = 0.5, + .gpuvm_min_page_size_bytes = 4096, + .hostvm_min_page_size_bytes = 4096, + .do_urgent_latency_adjustment = false, + .urgent_latency_adjustment_fabric_clock_component_us = 0, + .urgent_latency_adjustment_fabric_clock_reference_mhz = 0, +}; + void dcn31_calculate_wm_and_dlg_fp( struct dc *dc, struct dc_state *context, display_e2e_pipe_params_st *pipes, @@ -632,3 +776,88 @@ void dcn315_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_param else dml_init_instance(&dc->dml, &dcn3_15_soc, &dcn3_15_ip, DML_PROJECT_DCN31_FPGA); } + +void dcn316_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params) +{ + struct clk_limit_table *clk_table = &bw_params->clk_table; + struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES]; + unsigned int i, closest_clk_lvl; + int max_dispclk_mhz = 0, max_dppclk_mhz = 0; + int j; + + dc_assert_fp_enabled(); + + // Default clock levels are used for diags, which may lead to overclocking. + if (!IS_DIAG_DC(dc->ctx->dce_environment)) { + + dcn3_16_ip.max_num_otg = dc->res_pool->res_cap->num_timing_generator; + dcn3_16_ip.max_num_dpp = dc->res_pool->pipe_count; + dcn3_16_soc.num_chans = bw_params->num_channels; + + ASSERT(clk_table->num_entries); + + /* Prepass to find max clocks independent of voltage level. */ + for (i = 0; i < clk_table->num_entries; ++i) { + if (clk_table->entries[i].dispclk_mhz > max_dispclk_mhz) + max_dispclk_mhz = clk_table->entries[i].dispclk_mhz; + if (clk_table->entries[i].dppclk_mhz > max_dppclk_mhz) + max_dppclk_mhz = clk_table->entries[i].dppclk_mhz; + } + + for (i = 0; i < clk_table->num_entries; i++) { + /* loop backwards*/ + for (closest_clk_lvl = 0, j = dcn3_16_soc.num_states - 1; j >= 0; j--) { + if ((unsigned int) dcn3_16_soc.clock_limits[j].dcfclk_mhz <= clk_table->entries[i].dcfclk_mhz) { + closest_clk_lvl = j; + break; + } + } + // Ported from DCN315 + if (clk_table->num_entries == 1) { + /*smu gives one DPM level, let's take the highest one*/ + closest_clk_lvl = dcn3_16_soc.num_states - 1; + } + + clock_limits[i].state = i; + + /* Clocks dependent on voltage level. */ + clock_limits[i].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz; + if (clk_table->num_entries == 1 && + clock_limits[i].dcfclk_mhz < dcn3_16_soc.clock_limits[closest_clk_lvl].dcfclk_mhz) { + /*SMU fix not released yet*/ + clock_limits[i].dcfclk_mhz = dcn3_16_soc.clock_limits[closest_clk_lvl].dcfclk_mhz; + } + clock_limits[i].fabricclk_mhz = clk_table->entries[i].fclk_mhz; + clock_limits[i].socclk_mhz = clk_table->entries[i].socclk_mhz; + clock_limits[i].dram_speed_mts = clk_table->entries[i].memclk_mhz * 2 * clk_table->entries[i].wck_ratio; + + /* Clocks independent of voltage level. */ + clock_limits[i].dispclk_mhz = max_dispclk_mhz ? max_dispclk_mhz : + dcn3_16_soc.clock_limits[closest_clk_lvl].dispclk_mhz; + + clock_limits[i].dppclk_mhz = max_dppclk_mhz ? max_dppclk_mhz : + dcn3_16_soc.clock_limits[closest_clk_lvl].dppclk_mhz; + + clock_limits[i].dram_bw_per_chan_gbps = dcn3_16_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps; + clock_limits[i].dscclk_mhz = dcn3_16_soc.clock_limits[closest_clk_lvl].dscclk_mhz; + clock_limits[i].dtbclk_mhz = dcn3_16_soc.clock_limits[closest_clk_lvl].dtbclk_mhz; + clock_limits[i].phyclk_d18_mhz = dcn3_16_soc.clock_limits[closest_clk_lvl].phyclk_d18_mhz; + clock_limits[i].phyclk_mhz = dcn3_16_soc.clock_limits[closest_clk_lvl].phyclk_mhz; + } + for (i = 0; i < clk_table->num_entries; i++) + dcn3_16_soc.clock_limits[i] = clock_limits[i]; + if (clk_table->num_entries) { + dcn3_16_soc.num_states = clk_table->num_entries; + } + } + + if (max_dispclk_mhz) { + dcn3_16_soc.dispclk_dppclk_vco_speed_mhz = max_dispclk_mhz * 2; + dc->dml.soc.dispclk_dppclk_vco_speed_mhz = max_dispclk_mhz * 2; + } + + if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) + dml_init_instance(&dc->dml, &dcn3_16_soc, &dcn3_16_ip, DML_PROJECT_DCN31); + else + dml_init_instance(&dc->dml, &dcn3_16_soc, &dcn3_16_ip, DML_PROJECT_DCN31_FPGA); +} diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h index b15b587cf8c4..24ac19c83687 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h @@ -29,6 +29,7 @@ #define DCN3_1_DEFAULT_DET_SIZE 384 #define DCN3_15_DEFAULT_DET_SIZE 192 #define DCN3_15_MIN_COMPBUF_SIZE_KB 128 +#define DCN3_16_DEFAULT_DET_SIZE 192 void dcn31_calculate_wm_and_dlg_fp( struct dc *dc, struct dc_state *context, @@ -38,5 +39,6 @@ void dcn31_calculate_wm_and_dlg_fp( void dcn31_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params); void dcn315_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params); +void dcn316_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params); #endif /* __DCN31_FPU_H__*/ -- cgit v1.2.3 From 6e97c2f9689227b3fb6e2563127669f2146b5df5 Mon Sep 17 00:00:00 2001 From: Christian König Date: Mon, 14 Mar 2022 13:11:28 +0100 Subject: drm/amdgpu: move VM PDEs to idle after update MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Move the page tables to the idle list after updating the PDEs. We have gone back and forth with that a couple of times because of problems with the inter PD dependencies, but it should work now that we have the state handling cleanly separated. Signed-off-by: Christian König Acked-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 42 ++++++++-------------------------- 1 file changed, 10 insertions(+), 32 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index fc4563cf2828..eb9b86466336 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -1363,25 +1363,6 @@ static int amdgpu_vm_update_pde(struct amdgpu_vm_update_params *params, 1, 0, flags); } -/** - * amdgpu_vm_invalidate_pds - mark all PDs as invalid - * - * @adev: amdgpu_device pointer - * @vm: related vm - * - * Mark all PD level as invalid after an error. - */ -static void amdgpu_vm_invalidate_pds(struct amdgpu_device *adev, - struct amdgpu_vm *vm) -{ - struct amdgpu_vm_pt_cursor cursor; - struct amdgpu_vm_bo_base *entry; - - for_each_amdgpu_vm_pt_dfs_safe(adev, vm, NULL, cursor, entry) - if (entry->bo && !entry->moved) - amdgpu_vm_bo_relocated(entry); -} - /** * amdgpu_vm_update_pdes - make sure that all directories are valid * @@ -1398,6 +1379,7 @@ int amdgpu_vm_update_pdes(struct amdgpu_device *adev, struct amdgpu_vm *vm, bool immediate) { struct amdgpu_vm_update_params params; + struct amdgpu_vm_bo_base *entry; int r, idx; if (list_empty(&vm->relocated)) @@ -1413,16 +1395,9 @@ int amdgpu_vm_update_pdes(struct amdgpu_device *adev, r = vm->update_funcs->prepare(¶ms, NULL, AMDGPU_SYNC_EXPLICIT); if (r) - goto exit; - - while (!list_empty(&vm->relocated)) { - struct amdgpu_vm_bo_base *entry; - - entry = list_first_entry(&vm->relocated, - struct amdgpu_vm_bo_base, - vm_status); - amdgpu_vm_bo_idle(entry); + goto error; + list_for_each_entry(entry, &vm->relocated, vm_status) { r = amdgpu_vm_update_pde(¶ms, vm, entry); if (r) goto error; @@ -1431,12 +1406,15 @@ int amdgpu_vm_update_pdes(struct amdgpu_device *adev, r = vm->update_funcs->commit(¶ms, &vm->last_update); if (r) goto error; - drm_dev_exit(idx); - return 0; + + while (!list_empty(&vm->relocated)) { + entry = list_first_entry(&vm->relocated, + struct amdgpu_vm_bo_base, + vm_status); + amdgpu_vm_bo_idle(entry); + } error: - amdgpu_vm_invalidate_pds(adev, vm); -exit: drm_dev_exit(idx); return r; } -- cgit v1.2.3 From 184a69ca4d4125eee07f853cde45ee19dd21c437 Mon Sep 17 00:00:00 2001 From: Christian König Date: Tue, 15 Mar 2022 08:51:24 +0100 Subject: drm/amdgpu: separate VM PT handling into amdgpu_vm_pt.c MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Separate the VM page table backend operations from the state machine since the amdgpu_vm.c file is becoming to complex. The allocating, freeing and updating page tables and page directories can easily be moved into a separate file. While at it cleanup everything checkpatch.pl reported and rename the functions a bit to make more clear that they belong together. Signed-off-by: Christian König Acked-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/Makefile | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 953 +---------------------------- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 16 + drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c | 979 ++++++++++++++++++++++++++++++ 4 files changed, 1006 insertions(+), 944 deletions(-) create mode 100644 drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile index 40e2c6e2df79..7d7af43a258f 100644 --- a/drivers/gpu/drm/amd/amdgpu/Makefile +++ b/drivers/gpu/drm/amd/amdgpu/Makefile @@ -49,7 +49,7 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \ amdgpu_cs.o amdgpu_bios.o amdgpu_benchmark.o \ atombios_dp.o amdgpu_afmt.o amdgpu_trace_points.o \ atombios_encoders.o amdgpu_sa.o atombios_i2c.o \ - amdgpu_dma_buf.o amdgpu_vm.o amdgpu_ib.o amdgpu_pll.o \ + amdgpu_dma_buf.o amdgpu_vm.o amdgpu_vm_pt.o amdgpu_ib.o amdgpu_pll.o \ amdgpu_ucode.o amdgpu_bo_list.o amdgpu_ctx.o amdgpu_sync.o \ amdgpu_gtt_mgr.o amdgpu_preempt_mgr.o amdgpu_vram_mgr.o amdgpu_virt.o \ amdgpu_atomfirmware.o amdgpu_vf_error.o amdgpu_sched.o \ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index eb9b86466336..accebbac2f29 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -154,108 +154,6 @@ static inline void amdgpu_vm_eviction_unlock(struct amdgpu_vm *vm) mutex_unlock(&vm->eviction_lock); } -/** - * amdgpu_vm_level_shift - return the addr shift for each level - * - * @adev: amdgpu_device pointer - * @level: VMPT level - * - * Returns: - * The number of bits the pfn needs to be right shifted for a level. - */ -static unsigned amdgpu_vm_level_shift(struct amdgpu_device *adev, - unsigned level) -{ - switch (level) { - case AMDGPU_VM_PDB2: - case AMDGPU_VM_PDB1: - case AMDGPU_VM_PDB0: - return 9 * (AMDGPU_VM_PDB0 - level) + - adev->vm_manager.block_size; - case AMDGPU_VM_PTB: - return 0; - default: - return ~0; - } -} - -/** - * amdgpu_vm_num_entries - return the number of entries in a PD/PT - * - * @adev: amdgpu_device pointer - * @level: VMPT level - * - * Returns: - * The number of entries in a page directory or page table. - */ -static unsigned amdgpu_vm_num_entries(struct amdgpu_device *adev, - unsigned level) -{ - unsigned shift = amdgpu_vm_level_shift(adev, - adev->vm_manager.root_level); - - if (level == adev->vm_manager.root_level) - /* For the root directory */ - return round_up(adev->vm_manager.max_pfn, 1ULL << shift) - >> shift; - else if (level != AMDGPU_VM_PTB) - /* Everything in between */ - return 512; - else - /* For the page tables on the leaves */ - return AMDGPU_VM_PTE_COUNT(adev); -} - -/** - * amdgpu_vm_num_ats_entries - return the number of ATS entries in the root PD - * - * @adev: amdgpu_device pointer - * - * Returns: - * The number of entries in the root page directory which needs the ATS setting. - */ -static unsigned amdgpu_vm_num_ats_entries(struct amdgpu_device *adev) -{ - unsigned shift; - - shift = amdgpu_vm_level_shift(adev, adev->vm_manager.root_level); - return AMDGPU_GMC_HOLE_START >> (shift + AMDGPU_GPU_PAGE_SHIFT); -} - -/** - * amdgpu_vm_entries_mask - the mask to get the entry number of a PD/PT - * - * @adev: amdgpu_device pointer - * @level: VMPT level - * - * Returns: - * The mask to extract the entry number of a PD/PT from an address. - */ -static uint32_t amdgpu_vm_entries_mask(struct amdgpu_device *adev, - unsigned int level) -{ - if (level <= adev->vm_manager.root_level) - return 0xffffffff; - else if (level != AMDGPU_VM_PTB) - return 0x1ff; - else - return AMDGPU_VM_PTE_COUNT(adev) - 1; -} - -/** - * amdgpu_vm_bo_size - returns the size of the BOs in bytes - * - * @adev: amdgpu_device pointer - * @level: VMPT level - * - * Returns: - * The size of the BO for a page directory or page table in bytes. - */ -static unsigned amdgpu_vm_bo_size(struct amdgpu_device *adev, unsigned level) -{ - return AMDGPU_GPU_PAGE_ALIGN(amdgpu_vm_num_entries(adev, level) * 8); -} - /** * amdgpu_vm_bo_evicted - vm_bo is evicted * @@ -358,9 +256,8 @@ static void amdgpu_vm_bo_done(struct amdgpu_vm_bo_base *vm_bo) * Initialize a bo_va_base structure and add it to the appropriate lists * */ -static void amdgpu_vm_bo_base_init(struct amdgpu_vm_bo_base *base, - struct amdgpu_vm *vm, - struct amdgpu_bo *bo) +void amdgpu_vm_bo_base_init(struct amdgpu_vm_bo_base *base, + struct amdgpu_vm *vm, struct amdgpu_bo *bo) { base->vm = vm; base->bo = bo; @@ -395,228 +292,6 @@ static void amdgpu_vm_bo_base_init(struct amdgpu_vm_bo_base *base, amdgpu_vm_bo_evicted(base); } -/** - * amdgpu_vm_pt_parent - get the parent page directory - * - * @pt: child page table - * - * Helper to get the parent entry for the child page table. NULL if we are at - * the root page directory. - */ -static struct amdgpu_vm_bo_base *amdgpu_vm_pt_parent(struct amdgpu_vm_bo_base *pt) -{ - struct amdgpu_bo *parent = pt->bo->parent; - - if (!parent) - return NULL; - - return parent->vm_bo; -} - -/* - * amdgpu_vm_pt_cursor - state for for_each_amdgpu_vm_pt - */ -struct amdgpu_vm_pt_cursor { - uint64_t pfn; - struct amdgpu_vm_bo_base *parent; - struct amdgpu_vm_bo_base *entry; - unsigned level; -}; - -/** - * amdgpu_vm_pt_start - start PD/PT walk - * - * @adev: amdgpu_device pointer - * @vm: amdgpu_vm structure - * @start: start address of the walk - * @cursor: state to initialize - * - * Initialize a amdgpu_vm_pt_cursor to start a walk. - */ -static void amdgpu_vm_pt_start(struct amdgpu_device *adev, - struct amdgpu_vm *vm, uint64_t start, - struct amdgpu_vm_pt_cursor *cursor) -{ - cursor->pfn = start; - cursor->parent = NULL; - cursor->entry = &vm->root; - cursor->level = adev->vm_manager.root_level; -} - -/** - * amdgpu_vm_pt_descendant - go to child node - * - * @adev: amdgpu_device pointer - * @cursor: current state - * - * Walk to the child node of the current node. - * Returns: - * True if the walk was possible, false otherwise. - */ -static bool amdgpu_vm_pt_descendant(struct amdgpu_device *adev, - struct amdgpu_vm_pt_cursor *cursor) -{ - unsigned mask, shift, idx; - - if ((cursor->level == AMDGPU_VM_PTB) || !cursor->entry || - !cursor->entry->bo) - return false; - - mask = amdgpu_vm_entries_mask(adev, cursor->level); - shift = amdgpu_vm_level_shift(adev, cursor->level); - - ++cursor->level; - idx = (cursor->pfn >> shift) & mask; - cursor->parent = cursor->entry; - cursor->entry = &to_amdgpu_bo_vm(cursor->entry->bo)->entries[idx]; - return true; -} - -/** - * amdgpu_vm_pt_sibling - go to sibling node - * - * @adev: amdgpu_device pointer - * @cursor: current state - * - * Walk to the sibling node of the current node. - * Returns: - * True if the walk was possible, false otherwise. - */ -static bool amdgpu_vm_pt_sibling(struct amdgpu_device *adev, - struct amdgpu_vm_pt_cursor *cursor) -{ - unsigned shift, num_entries; - - /* Root doesn't have a sibling */ - if (!cursor->parent) - return false; - - /* Go to our parents and see if we got a sibling */ - shift = amdgpu_vm_level_shift(adev, cursor->level - 1); - num_entries = amdgpu_vm_num_entries(adev, cursor->level - 1); - - if (cursor->entry == &to_amdgpu_bo_vm(cursor->parent->bo)->entries[num_entries - 1]) - return false; - - cursor->pfn += 1ULL << shift; - cursor->pfn &= ~((1ULL << shift) - 1); - ++cursor->entry; - return true; -} - -/** - * amdgpu_vm_pt_ancestor - go to parent node - * - * @cursor: current state - * - * Walk to the parent node of the current node. - * Returns: - * True if the walk was possible, false otherwise. - */ -static bool amdgpu_vm_pt_ancestor(struct amdgpu_vm_pt_cursor *cursor) -{ - if (!cursor->parent) - return false; - - --cursor->level; - cursor->entry = cursor->parent; - cursor->parent = amdgpu_vm_pt_parent(cursor->parent); - return true; -} - -/** - * amdgpu_vm_pt_next - get next PD/PT in hieratchy - * - * @adev: amdgpu_device pointer - * @cursor: current state - * - * Walk the PD/PT tree to the next node. - */ -static void amdgpu_vm_pt_next(struct amdgpu_device *adev, - struct amdgpu_vm_pt_cursor *cursor) -{ - /* First try a newborn child */ - if (amdgpu_vm_pt_descendant(adev, cursor)) - return; - - /* If that didn't worked try to find a sibling */ - while (!amdgpu_vm_pt_sibling(adev, cursor)) { - /* No sibling, go to our parents and grandparents */ - if (!amdgpu_vm_pt_ancestor(cursor)) { - cursor->pfn = ~0ll; - return; - } - } -} - -/** - * amdgpu_vm_pt_first_dfs - start a deep first search - * - * @adev: amdgpu_device structure - * @vm: amdgpu_vm structure - * @start: optional cursor to start with - * @cursor: state to initialize - * - * Starts a deep first traversal of the PD/PT tree. - */ -static void amdgpu_vm_pt_first_dfs(struct amdgpu_device *adev, - struct amdgpu_vm *vm, - struct amdgpu_vm_pt_cursor *start, - struct amdgpu_vm_pt_cursor *cursor) -{ - if (start) - *cursor = *start; - else - amdgpu_vm_pt_start(adev, vm, 0, cursor); - while (amdgpu_vm_pt_descendant(adev, cursor)); -} - -/** - * amdgpu_vm_pt_continue_dfs - check if the deep first search should continue - * - * @start: starting point for the search - * @entry: current entry - * - * Returns: - * True when the search should continue, false otherwise. - */ -static bool amdgpu_vm_pt_continue_dfs(struct amdgpu_vm_pt_cursor *start, - struct amdgpu_vm_bo_base *entry) -{ - return entry && (!start || entry != start->entry); -} - -/** - * amdgpu_vm_pt_next_dfs - get the next node for a deep first search - * - * @adev: amdgpu_device structure - * @cursor: current state - * - * Move the cursor to the next node in a deep first search. - */ -static void amdgpu_vm_pt_next_dfs(struct amdgpu_device *adev, - struct amdgpu_vm_pt_cursor *cursor) -{ - if (!cursor->entry) - return; - - if (!cursor->parent) - cursor->entry = NULL; - else if (amdgpu_vm_pt_sibling(adev, cursor)) - while (amdgpu_vm_pt_descendant(adev, cursor)); - else - amdgpu_vm_pt_ancestor(cursor); -} - -/* - * for_each_amdgpu_vm_pt_dfs_safe - safe deep first search of all PDs/PTs - */ -#define for_each_amdgpu_vm_pt_dfs_safe(adev, vm, start, cursor, entry) \ - for (amdgpu_vm_pt_first_dfs((adev), (vm), (start), &(cursor)), \ - (entry) = (cursor).entry, amdgpu_vm_pt_next_dfs((adev), &(cursor));\ - amdgpu_vm_pt_continue_dfs((start), (entry)); \ - (entry) = (cursor).entry, amdgpu_vm_pt_next_dfs((adev), &(cursor))) - /** * amdgpu_vm_get_pd_bo - add the VM PD to a validation list * @@ -783,312 +458,6 @@ bool amdgpu_vm_ready(struct amdgpu_vm *vm) return ret && list_empty(&vm->evicted); } -/** - * amdgpu_vm_clear_bo - initially clear the PDs/PTs - * - * @adev: amdgpu_device pointer - * @vm: VM to clear BO from - * @vmbo: BO to clear - * @immediate: use an immediate update - * - * Root PD needs to be reserved when calling this. - * - * Returns: - * 0 on success, errno otherwise. - */ -static int amdgpu_vm_clear_bo(struct amdgpu_device *adev, - struct amdgpu_vm *vm, - struct amdgpu_bo_vm *vmbo, - bool immediate) -{ - struct ttm_operation_ctx ctx = { true, false }; - unsigned level = adev->vm_manager.root_level; - struct amdgpu_vm_update_params params; - struct amdgpu_bo *ancestor = &vmbo->bo; - struct amdgpu_bo *bo = &vmbo->bo; - unsigned entries, ats_entries; - uint64_t addr; - int r, idx; - - /* Figure out our place in the hierarchy */ - if (ancestor->parent) { - ++level; - while (ancestor->parent->parent) { - ++level; - ancestor = ancestor->parent; - } - } - - entries = amdgpu_bo_size(bo) / 8; - if (!vm->pte_support_ats) { - ats_entries = 0; - - } else if (!bo->parent) { - ats_entries = amdgpu_vm_num_ats_entries(adev); - ats_entries = min(ats_entries, entries); - entries -= ats_entries; - - } else { - struct amdgpu_vm_bo_base *pt; - - pt = ancestor->vm_bo; - ats_entries = amdgpu_vm_num_ats_entries(adev); - if ((pt - to_amdgpu_bo_vm(vm->root.bo)->entries) >= ats_entries) { - ats_entries = 0; - } else { - ats_entries = entries; - entries = 0; - } - } - - r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); - if (r) - return r; - - if (vmbo->shadow) { - struct amdgpu_bo *shadow = vmbo->shadow; - - r = ttm_bo_validate(&shadow->tbo, &shadow->placement, &ctx); - if (r) - return r; - } - - if (!drm_dev_enter(adev_to_drm(adev), &idx)) - return -ENODEV; - - r = vm->update_funcs->map_table(vmbo); - if (r) - goto exit; - - memset(¶ms, 0, sizeof(params)); - params.adev = adev; - params.vm = vm; - params.immediate = immediate; - - r = vm->update_funcs->prepare(¶ms, NULL, AMDGPU_SYNC_EXPLICIT); - if (r) - goto exit; - - addr = 0; - if (ats_entries) { - uint64_t value = 0, flags; - - flags = AMDGPU_PTE_DEFAULT_ATC; - if (level != AMDGPU_VM_PTB) { - /* Handle leaf PDEs as PTEs */ - flags |= AMDGPU_PDE_PTE; - amdgpu_gmc_get_vm_pde(adev, level, &value, &flags); - } - - r = vm->update_funcs->update(¶ms, vmbo, addr, 0, ats_entries, - value, flags); - if (r) - goto exit; - - addr += ats_entries * 8; - } - - if (entries) { - uint64_t value = 0, flags = 0; - - if (adev->asic_type >= CHIP_VEGA10) { - if (level != AMDGPU_VM_PTB) { - /* Handle leaf PDEs as PTEs */ - flags |= AMDGPU_PDE_PTE; - amdgpu_gmc_get_vm_pde(adev, level, - &value, &flags); - } else { - /* Workaround for fault priority problem on GMC9 */ - flags = AMDGPU_PTE_EXECUTABLE; - } - } - - r = vm->update_funcs->update(¶ms, vmbo, addr, 0, entries, - value, flags); - if (r) - goto exit; - } - - r = vm->update_funcs->commit(¶ms, NULL); -exit: - drm_dev_exit(idx); - return r; -} - -/** - * amdgpu_vm_pt_create - create bo for PD/PT - * - * @adev: amdgpu_device pointer - * @vm: requesting vm - * @level: the page table level - * @immediate: use a immediate update - * @vmbo: pointer to the buffer object pointer - */ -static int amdgpu_vm_pt_create(struct amdgpu_device *adev, - struct amdgpu_vm *vm, - int level, bool immediate, - struct amdgpu_bo_vm **vmbo) -{ - struct amdgpu_bo_param bp; - struct amdgpu_bo *bo; - struct dma_resv *resv; - unsigned int num_entries; - int r; - - memset(&bp, 0, sizeof(bp)); - - bp.size = amdgpu_vm_bo_size(adev, level); - bp.byte_align = AMDGPU_GPU_PAGE_SIZE; - bp.domain = AMDGPU_GEM_DOMAIN_VRAM; - bp.domain = amdgpu_bo_get_preferred_domain(adev, bp.domain); - bp.flags = AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS | - AMDGPU_GEM_CREATE_CPU_GTT_USWC; - - if (level < AMDGPU_VM_PTB) - num_entries = amdgpu_vm_num_entries(adev, level); - else - num_entries = 0; - - bp.bo_ptr_size = struct_size((*vmbo), entries, num_entries); - - if (vm->use_cpu_for_update) - bp.flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; - - bp.type = ttm_bo_type_kernel; - bp.no_wait_gpu = immediate; - if (vm->root.bo) - bp.resv = vm->root.bo->tbo.base.resv; - - r = amdgpu_bo_create_vm(adev, &bp, vmbo); - if (r) - return r; - - bo = &(*vmbo)->bo; - if (vm->is_compute_context || (adev->flags & AMD_IS_APU)) { - (*vmbo)->shadow = NULL; - return 0; - } - - if (!bp.resv) - WARN_ON(dma_resv_lock(bo->tbo.base.resv, - NULL)); - resv = bp.resv; - memset(&bp, 0, sizeof(bp)); - bp.size = amdgpu_vm_bo_size(adev, level); - bp.domain = AMDGPU_GEM_DOMAIN_GTT; - bp.flags = AMDGPU_GEM_CREATE_CPU_GTT_USWC; - bp.type = ttm_bo_type_kernel; - bp.resv = bo->tbo.base.resv; - bp.bo_ptr_size = sizeof(struct amdgpu_bo); - - r = amdgpu_bo_create(adev, &bp, &(*vmbo)->shadow); - - if (!resv) - dma_resv_unlock(bo->tbo.base.resv); - - if (r) { - amdgpu_bo_unref(&bo); - return r; - } - - (*vmbo)->shadow->parent = amdgpu_bo_ref(bo); - amdgpu_bo_add_to_shadow_list(*vmbo); - - return 0; -} - -/** - * amdgpu_vm_alloc_pts - Allocate a specific page table - * - * @adev: amdgpu_device pointer - * @vm: VM to allocate page tables for - * @cursor: Which page table to allocate - * @immediate: use an immediate update - * - * Make sure a specific page table or directory is allocated. - * - * Returns: - * 1 if page table needed to be allocated, 0 if page table was already - * allocated, negative errno if an error occurred. - */ -static int amdgpu_vm_alloc_pts(struct amdgpu_device *adev, - struct amdgpu_vm *vm, - struct amdgpu_vm_pt_cursor *cursor, - bool immediate) -{ - struct amdgpu_vm_bo_base *entry = cursor->entry; - struct amdgpu_bo *pt_bo; - struct amdgpu_bo_vm *pt; - int r; - - if (entry->bo) - return 0; - - r = amdgpu_vm_pt_create(adev, vm, cursor->level, immediate, &pt); - if (r) - return r; - - /* Keep a reference to the root directory to avoid - * freeing them up in the wrong order. - */ - pt_bo = &pt->bo; - pt_bo->parent = amdgpu_bo_ref(cursor->parent->bo); - amdgpu_vm_bo_base_init(entry, vm, pt_bo); - r = amdgpu_vm_clear_bo(adev, vm, pt, immediate); - if (r) - goto error_free_pt; - - return 0; - -error_free_pt: - amdgpu_bo_unref(&pt->shadow); - amdgpu_bo_unref(&pt_bo); - return r; -} - -/** - * amdgpu_vm_free_table - fre one PD/PT - * - * @entry: PDE to free - */ -static void amdgpu_vm_free_table(struct amdgpu_vm_bo_base *entry) -{ - struct amdgpu_bo *shadow; - - if (!entry->bo) - return; - shadow = amdgpu_bo_shadowed(entry->bo); - entry->bo->vm_bo = NULL; - list_del(&entry->vm_status); - amdgpu_bo_unref(&shadow); - amdgpu_bo_unref(&entry->bo); -} - -/** - * amdgpu_vm_free_pts - free PD/PT levels - * - * @adev: amdgpu device structure - * @vm: amdgpu vm structure - * @start: optional cursor where to start freeing PDs/PTs - * - * Free the page directory or page table level and all sub levels. - */ -static void amdgpu_vm_free_pts(struct amdgpu_device *adev, - struct amdgpu_vm *vm, - struct amdgpu_vm_pt_cursor *start) -{ - struct amdgpu_vm_pt_cursor cursor; - struct amdgpu_vm_bo_base *entry; - - vm->bulk_moveable = false; - - for_each_amdgpu_vm_pt_dfs_safe(adev, vm, start, cursor, entry) - amdgpu_vm_free_table(entry); - - if (start) - amdgpu_vm_free_table(start->entry); -} - /** * amdgpu_vm_check_compute_bug - check whether asic has compute vm bug * @@ -1335,34 +704,6 @@ uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr) return result; } -/** - * amdgpu_vm_update_pde - update a single level in the hierarchy - * - * @params: parameters for the update - * @vm: requested vm - * @entry: entry to update - * - * Makes sure the requested entry in parent is up to date. - */ -static int amdgpu_vm_update_pde(struct amdgpu_vm_update_params *params, - struct amdgpu_vm *vm, - struct amdgpu_vm_bo_base *entry) -{ - struct amdgpu_vm_bo_base *parent = amdgpu_vm_pt_parent(entry); - struct amdgpu_bo *bo = parent->bo, *pbo; - uint64_t pde, pt, flags; - unsigned level; - - for (level = 0, pbo = bo->parent; pbo; ++level) - pbo = pbo->parent; - - level += params->adev->vm_manager.root_level; - amdgpu_gmc_get_pde_for_bo(entry->bo, level, &pt, &flags); - pde = (entry - to_amdgpu_bo_vm(parent->bo)->entries) * 8; - return vm->update_funcs->update(params, to_amdgpu_bo_vm(bo), pde, pt, - 1, 0, flags); -} - /** * amdgpu_vm_update_pdes - make sure that all directories are valid * @@ -1398,7 +739,7 @@ int amdgpu_vm_update_pdes(struct amdgpu_device *adev, goto error; list_for_each_entry(entry, &vm->relocated, vm_status) { - r = amdgpu_vm_update_pde(¶ms, vm, entry); + r = amdgpu_vm_pde_update(¶ms, entry); if (r) goto error; } @@ -1419,252 +760,6 @@ error: return r; } -/* - * amdgpu_vm_update_flags - figure out flags for PTE updates - * - * Make sure to set the right flags for the PTEs at the desired level. - */ -static void amdgpu_vm_update_flags(struct amdgpu_vm_update_params *params, - struct amdgpu_bo_vm *pt, unsigned int level, - uint64_t pe, uint64_t addr, - unsigned int count, uint32_t incr, - uint64_t flags) - -{ - if (level != AMDGPU_VM_PTB) { - flags |= AMDGPU_PDE_PTE; - amdgpu_gmc_get_vm_pde(params->adev, level, &addr, &flags); - - } else if (params->adev->asic_type >= CHIP_VEGA10 && - !(flags & AMDGPU_PTE_VALID) && - !(flags & AMDGPU_PTE_PRT)) { - - /* Workaround for fault priority problem on GMC9 */ - flags |= AMDGPU_PTE_EXECUTABLE; - } - - params->vm->update_funcs->update(params, pt, pe, addr, count, incr, - flags); -} - -/** - * amdgpu_vm_fragment - get fragment for PTEs - * - * @params: see amdgpu_vm_update_params definition - * @start: first PTE to handle - * @end: last PTE to handle - * @flags: hw mapping flags - * @frag: resulting fragment size - * @frag_end: end of this fragment - * - * Returns the first possible fragment for the start and end address. - */ -static void amdgpu_vm_fragment(struct amdgpu_vm_update_params *params, - uint64_t start, uint64_t end, uint64_t flags, - unsigned int *frag, uint64_t *frag_end) -{ - /** - * The MC L1 TLB supports variable sized pages, based on a fragment - * field in the PTE. When this field is set to a non-zero value, page - * granularity is increased from 4KB to (1 << (12 + frag)). The PTE - * flags are considered valid for all PTEs within the fragment range - * and corresponding mappings are assumed to be physically contiguous. - * - * The L1 TLB can store a single PTE for the whole fragment, - * significantly increasing the space available for translation - * caching. This leads to large improvements in throughput when the - * TLB is under pressure. - * - * The L2 TLB distributes small and large fragments into two - * asymmetric partitions. The large fragment cache is significantly - * larger. Thus, we try to use large fragments wherever possible. - * Userspace can support this by aligning virtual base address and - * allocation size to the fragment size. - * - * Starting with Vega10 the fragment size only controls the L1. The L2 - * is now directly feed with small/huge/giant pages from the walker. - */ - unsigned max_frag; - - if (params->adev->asic_type < CHIP_VEGA10) - max_frag = params->adev->vm_manager.fragment_size; - else - max_frag = 31; - - /* system pages are non continuously */ - if (params->pages_addr) { - *frag = 0; - *frag_end = end; - return; - } - - /* This intentionally wraps around if no bit is set */ - *frag = min((unsigned)ffs(start) - 1, (unsigned)fls64(end - start) - 1); - if (*frag >= max_frag) { - *frag = max_frag; - *frag_end = end & ~((1ULL << max_frag) - 1); - } else { - *frag_end = start + (1 << *frag); - } -} - -/** - * amdgpu_vm_update_ptes - make sure that page tables are valid - * - * @params: see amdgpu_vm_update_params definition - * @start: start of GPU address range - * @end: end of GPU address range - * @dst: destination address to map to, the next dst inside the function - * @flags: mapping flags - * - * Update the page tables in the range @start - @end. - * - * Returns: - * 0 for success, -EINVAL for failure. - */ -static int amdgpu_vm_update_ptes(struct amdgpu_vm_update_params *params, - uint64_t start, uint64_t end, - uint64_t dst, uint64_t flags) -{ - struct amdgpu_device *adev = params->adev; - struct amdgpu_vm_pt_cursor cursor; - uint64_t frag_start = start, frag_end; - unsigned int frag; - int r; - - /* figure out the initial fragment */ - amdgpu_vm_fragment(params, frag_start, end, flags, &frag, &frag_end); - - /* walk over the address space and update the PTs */ - amdgpu_vm_pt_start(adev, params->vm, start, &cursor); - while (cursor.pfn < end) { - unsigned shift, parent_shift, mask; - uint64_t incr, entry_end, pe_start; - struct amdgpu_bo *pt; - - if (!params->unlocked) { - /* make sure that the page tables covering the - * address range are actually allocated - */ - r = amdgpu_vm_alloc_pts(params->adev, params->vm, - &cursor, params->immediate); - if (r) - return r; - } - - shift = amdgpu_vm_level_shift(adev, cursor.level); - parent_shift = amdgpu_vm_level_shift(adev, cursor.level - 1); - if (params->unlocked) { - /* Unlocked updates are only allowed on the leaves */ - if (amdgpu_vm_pt_descendant(adev, &cursor)) - continue; - } else if (adev->asic_type < CHIP_VEGA10 && - (flags & AMDGPU_PTE_VALID)) { - /* No huge page support before GMC v9 */ - if (cursor.level != AMDGPU_VM_PTB) { - if (!amdgpu_vm_pt_descendant(adev, &cursor)) - return -ENOENT; - continue; - } - } else if (frag < shift) { - /* We can't use this level when the fragment size is - * smaller than the address shift. Go to the next - * child entry and try again. - */ - if (amdgpu_vm_pt_descendant(adev, &cursor)) - continue; - } else if (frag >= parent_shift) { - /* If the fragment size is even larger than the parent - * shift we should go up one level and check it again. - */ - if (!amdgpu_vm_pt_ancestor(&cursor)) - return -EINVAL; - continue; - } - - pt = cursor.entry->bo; - if (!pt) { - /* We need all PDs and PTs for mapping something, */ - if (flags & AMDGPU_PTE_VALID) - return -ENOENT; - - /* but unmapping something can happen at a higher - * level. - */ - if (!amdgpu_vm_pt_ancestor(&cursor)) - return -EINVAL; - - pt = cursor.entry->bo; - shift = parent_shift; - frag_end = max(frag_end, ALIGN(frag_start + 1, - 1ULL << shift)); - } - - /* Looks good so far, calculate parameters for the update */ - incr = (uint64_t)AMDGPU_GPU_PAGE_SIZE << shift; - mask = amdgpu_vm_entries_mask(adev, cursor.level); - pe_start = ((cursor.pfn >> shift) & mask) * 8; - entry_end = ((uint64_t)mask + 1) << shift; - entry_end += cursor.pfn & ~(entry_end - 1); - entry_end = min(entry_end, end); - - do { - struct amdgpu_vm *vm = params->vm; - uint64_t upd_end = min(entry_end, frag_end); - unsigned nptes = (upd_end - frag_start) >> shift; - uint64_t upd_flags = flags | AMDGPU_PTE_FRAG(frag); - - /* This can happen when we set higher level PDs to - * silent to stop fault floods. - */ - nptes = max(nptes, 1u); - - trace_amdgpu_vm_update_ptes(params, frag_start, upd_end, - min(nptes, 32u), dst, incr, upd_flags, - vm->task_info.pid, - vm->immediate.fence_context); - amdgpu_vm_update_flags(params, to_amdgpu_bo_vm(pt), - cursor.level, pe_start, dst, - nptes, incr, upd_flags); - - pe_start += nptes * 8; - dst += nptes * incr; - - frag_start = upd_end; - if (frag_start >= frag_end) { - /* figure out the next fragment */ - amdgpu_vm_fragment(params, frag_start, end, - flags, &frag, &frag_end); - if (frag < shift) - break; - } - } while (frag_start < entry_end); - - if (amdgpu_vm_pt_descendant(adev, &cursor)) { - /* Free all child entries. - * Update the tables with the flags and addresses and free up subsequent - * tables in the case of huge pages or freed up areas. - * This is the maximum you can free, because all other page tables are not - * completely covered by the range and so potentially still in use. - */ - while (cursor.pfn < frag_start) { - /* Make sure previous mapping is freed */ - if (cursor.entry->bo) { - params->table_freed = true; - amdgpu_vm_free_pts(adev, params->vm, &cursor); - } - amdgpu_vm_pt_next(adev, &cursor); - } - - } else if (frag >= shift) { - /* or just move on to the next on the same level. */ - amdgpu_vm_pt_next(adev, &cursor); - } - } - - return 0; -} - /** * amdgpu_vm_bo_update_mapping - update a mapping in the vm page table * @@ -1785,7 +880,7 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, } tmp = start + num_entries; - r = amdgpu_vm_update_ptes(¶ms, start, tmp, addr, flags); + r = amdgpu_vm_ptes_update(¶ms, start, tmp, addr, flags); if (r) goto error_unlock; @@ -2964,7 +2059,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm) amdgpu_vm_bo_base_init(&vm->root, vm, root_bo); - r = amdgpu_vm_clear_bo(adev, vm, root, false); + r = amdgpu_vm_pt_clear(adev, vm, root, false); if (r) goto error_unreserve; @@ -2992,34 +2087,6 @@ error_free_immediate: return r; } -/** - * amdgpu_vm_check_clean_reserved - check if a VM is clean - * - * @adev: amdgpu_device pointer - * @vm: the VM to check - * - * check all entries of the root PD, if any subsequent PDs are allocated, - * it means there are page table creating and filling, and is no a clean - * VM - * - * Returns: - * 0 if this VM is clean - */ -static int amdgpu_vm_check_clean_reserved(struct amdgpu_device *adev, - struct amdgpu_vm *vm) -{ - enum amdgpu_vm_level root = adev->vm_manager.root_level; - unsigned int entries = amdgpu_vm_num_entries(adev, root); - unsigned int i = 0; - - for (i = 0; i < entries; i++) { - if (to_amdgpu_bo_vm(vm->root.bo)->entries[i].bo) - return -EINVAL; - } - - return 0; -} - /** * amdgpu_vm_make_compute - Turn a GFX VM into a compute VM * @@ -3049,17 +2116,17 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm) return r; /* Sanity checks */ - r = amdgpu_vm_check_clean_reserved(adev, vm); - if (r) + if (!amdgpu_vm_pt_is_root_clean(adev, vm)) { + r = -EINVAL; goto unreserve_bo; + } /* Check if PD needs to be reinitialized and do it before * changing any other state, in case it fails. */ if (pte_support_ats != vm->pte_support_ats) { vm->pte_support_ats = pte_support_ats; - r = amdgpu_vm_clear_bo(adev, vm, - to_amdgpu_bo_vm(vm->root.bo), + r = amdgpu_vm_pt_clear(adev, vm, to_amdgpu_bo_vm(vm->root.bo), false); if (r) goto unreserve_bo; @@ -3147,7 +2214,7 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) amdgpu_vm_free_mapping(adev, vm, mapping, NULL); } - amdgpu_vm_free_pts(adev, vm, NULL); + amdgpu_vm_pt_free_root(adev, vm); amdgpu_bo_unreserve(root); amdgpu_bo_unref(&root); WARN_ON(vm->root.bo); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index a40a6a993bb0..1731681914f5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -397,6 +397,8 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev, struct dma_fence **fence); int amdgpu_vm_handle_moved(struct amdgpu_device *adev, struct amdgpu_vm *vm); +void amdgpu_vm_bo_base_init(struct amdgpu_vm_bo_base *base, + struct amdgpu_vm *vm, struct amdgpu_bo *bo); int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, struct amdgpu_device *bo_adev, struct amdgpu_vm *vm, bool immediate, @@ -458,6 +460,20 @@ void amdgpu_vm_del_from_lru_notify(struct ttm_buffer_object *bo); void amdgpu_vm_get_memory(struct amdgpu_vm *vm, uint64_t *vram_mem, uint64_t *gtt_mem, uint64_t *cpu_mem); +int amdgpu_vm_pt_clear(struct amdgpu_device *adev, struct amdgpu_vm *vm, + struct amdgpu_bo_vm *vmbo, bool immediate); +int amdgpu_vm_pt_create(struct amdgpu_device *adev, struct amdgpu_vm *vm, + int level, bool immediate, struct amdgpu_bo_vm **vmbo); +void amdgpu_vm_pt_free_root(struct amdgpu_device *adev, struct amdgpu_vm *vm); +bool amdgpu_vm_pt_is_root_clean(struct amdgpu_device *adev, + struct amdgpu_vm *vm); + +int amdgpu_vm_pde_update(struct amdgpu_vm_update_params *params, + struct amdgpu_vm_bo_base *entry); +int amdgpu_vm_ptes_update(struct amdgpu_vm_update_params *params, + uint64_t start, uint64_t end, + uint64_t dst, uint64_t flags); + #if defined(CONFIG_DEBUG_FS) void amdgpu_debugfs_vm_bo_info(struct amdgpu_vm *vm, struct seq_file *m); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c new file mode 100644 index 000000000000..a821ada5f8ca --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c @@ -0,0 +1,979 @@ +// SPDX-License-Identifier: GPL-2.0 OR MIT +/* + * Copyright 2022 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + */ + +#include + +#include "amdgpu.h" +#include "amdgpu_trace.h" +#include "amdgpu_vm.h" + +/* + * amdgpu_vm_pt_cursor - state for for_each_amdgpu_vm_pt + */ +struct amdgpu_vm_pt_cursor { + uint64_t pfn; + struct amdgpu_vm_bo_base *parent; + struct amdgpu_vm_bo_base *entry; + unsigned int level; +}; + +/** + * amdgpu_vm_pt_level_shift - return the addr shift for each level + * + * @adev: amdgpu_device pointer + * @level: VMPT level + * + * Returns: + * The number of bits the pfn needs to be right shifted for a level. + */ +static unsigned int amdgpu_vm_pt_level_shift(struct amdgpu_device *adev, + unsigned int level) +{ + switch (level) { + case AMDGPU_VM_PDB2: + case AMDGPU_VM_PDB1: + case AMDGPU_VM_PDB0: + return 9 * (AMDGPU_VM_PDB0 - level) + + adev->vm_manager.block_size; + case AMDGPU_VM_PTB: + return 0; + default: + return ~0; + } +} + +/** + * amdgpu_vm_pt_num_entries - return the number of entries in a PD/PT + * + * @adev: amdgpu_device pointer + * @level: VMPT level + * + * Returns: + * The number of entries in a page directory or page table. + */ +static unsigned int amdgpu_vm_pt_num_entries(struct amdgpu_device *adev, + unsigned int level) +{ + unsigned int shift; + + shift = amdgpu_vm_pt_level_shift(adev, adev->vm_manager.root_level); + if (level == adev->vm_manager.root_level) + /* For the root directory */ + return round_up(adev->vm_manager.max_pfn, 1ULL << shift) + >> shift; + else if (level != AMDGPU_VM_PTB) + /* Everything in between */ + return 512; + + /* For the page tables on the leaves */ + return AMDGPU_VM_PTE_COUNT(adev); +} + +/** + * amdgpu_vm_pt_num_ats_entries - return the number of ATS entries in the root PD + * + * @adev: amdgpu_device pointer + * + * Returns: + * The number of entries in the root page directory which needs the ATS setting. + */ +static unsigned int amdgpu_vm_pt_num_ats_entries(struct amdgpu_device *adev) +{ + unsigned int shift; + + shift = amdgpu_vm_pt_level_shift(adev, adev->vm_manager.root_level); + return AMDGPU_GMC_HOLE_START >> (shift + AMDGPU_GPU_PAGE_SHIFT); +} + +/** + * amdgpu_vm_pt_entries_mask - the mask to get the entry number of a PD/PT + * + * @adev: amdgpu_device pointer + * @level: VMPT level + * + * Returns: + * The mask to extract the entry number of a PD/PT from an address. + */ +static uint32_t amdgpu_vm_pt_entries_mask(struct amdgpu_device *adev, + unsigned int level) +{ + if (level <= adev->vm_manager.root_level) + return 0xffffffff; + else if (level != AMDGPU_VM_PTB) + return 0x1ff; + else + return AMDGPU_VM_PTE_COUNT(adev) - 1; +} + +/** + * amdgpu_vm_pt_size - returns the size of the page table in bytes + * + * @adev: amdgpu_device pointer + * @level: VMPT level + * + * Returns: + * The size of the BO for a page directory or page table in bytes. + */ +static unsigned int amdgpu_vm_pt_size(struct amdgpu_device *adev, + unsigned int level) +{ + return AMDGPU_GPU_PAGE_ALIGN(amdgpu_vm_pt_num_entries(adev, level) * 8); +} + +/** + * amdgpu_vm_pt_parent - get the parent page directory + * + * @pt: child page table + * + * Helper to get the parent entry for the child page table. NULL if we are at + * the root page directory. + */ +static struct amdgpu_vm_bo_base * +amdgpu_vm_pt_parent(struct amdgpu_vm_bo_base *pt) +{ + struct amdgpu_bo *parent = pt->bo->parent; + + if (!parent) + return NULL; + + return parent->vm_bo; +} + +/** + * amdgpu_vm_pt_start - start PD/PT walk + * + * @adev: amdgpu_device pointer + * @vm: amdgpu_vm structure + * @start: start address of the walk + * @cursor: state to initialize + * + * Initialize a amdgpu_vm_pt_cursor to start a walk. + */ +static void amdgpu_vm_pt_start(struct amdgpu_device *adev, + struct amdgpu_vm *vm, uint64_t start, + struct amdgpu_vm_pt_cursor *cursor) +{ + cursor->pfn = start; + cursor->parent = NULL; + cursor->entry = &vm->root; + cursor->level = adev->vm_manager.root_level; +} + +/** + * amdgpu_vm_pt_descendant - go to child node + * + * @adev: amdgpu_device pointer + * @cursor: current state + * + * Walk to the child node of the current node. + * Returns: + * True if the walk was possible, false otherwise. + */ +static bool amdgpu_vm_pt_descendant(struct amdgpu_device *adev, + struct amdgpu_vm_pt_cursor *cursor) +{ + unsigned int mask, shift, idx; + + if ((cursor->level == AMDGPU_VM_PTB) || !cursor->entry || + !cursor->entry->bo) + return false; + + mask = amdgpu_vm_pt_entries_mask(adev, cursor->level); + shift = amdgpu_vm_pt_level_shift(adev, cursor->level); + + ++cursor->level; + idx = (cursor->pfn >> shift) & mask; + cursor->parent = cursor->entry; + cursor->entry = &to_amdgpu_bo_vm(cursor->entry->bo)->entries[idx]; + return true; +} + +/** + * amdgpu_vm_pt_sibling - go to sibling node + * + * @adev: amdgpu_device pointer + * @cursor: current state + * + * Walk to the sibling node of the current node. + * Returns: + * True if the walk was possible, false otherwise. + */ +static bool amdgpu_vm_pt_sibling(struct amdgpu_device *adev, + struct amdgpu_vm_pt_cursor *cursor) +{ + + unsigned int shift, num_entries; + struct amdgpu_bo_vm *parent; + + /* Root doesn't have a sibling */ + if (!cursor->parent) + return false; + + /* Go to our parents and see if we got a sibling */ + shift = amdgpu_vm_pt_level_shift(adev, cursor->level - 1); + num_entries = amdgpu_vm_pt_num_entries(adev, cursor->level - 1); + parent = to_amdgpu_bo_vm(cursor->parent->bo); + + if (cursor->entry == &parent->entries[num_entries - 1]) + return false; + + cursor->pfn += 1ULL << shift; + cursor->pfn &= ~((1ULL << shift) - 1); + ++cursor->entry; + return true; +} + +/** + * amdgpu_vm_pt_ancestor - go to parent node + * + * @cursor: current state + * + * Walk to the parent node of the current node. + * Returns: + * True if the walk was possible, false otherwise. + */ +static bool amdgpu_vm_pt_ancestor(struct amdgpu_vm_pt_cursor *cursor) +{ + if (!cursor->parent) + return false; + + --cursor->level; + cursor->entry = cursor->parent; + cursor->parent = amdgpu_vm_pt_parent(cursor->parent); + return true; +} + +/** + * amdgpu_vm_pt_next - get next PD/PT in hieratchy + * + * @adev: amdgpu_device pointer + * @cursor: current state + * + * Walk the PD/PT tree to the next node. + */ +static void amdgpu_vm_pt_next(struct amdgpu_device *adev, + struct amdgpu_vm_pt_cursor *cursor) +{ + /* First try a newborn child */ + if (amdgpu_vm_pt_descendant(adev, cursor)) + return; + + /* If that didn't worked try to find a sibling */ + while (!amdgpu_vm_pt_sibling(adev, cursor)) { + /* No sibling, go to our parents and grandparents */ + if (!amdgpu_vm_pt_ancestor(cursor)) { + cursor->pfn = ~0ll; + return; + } + } +} + +/** + * amdgpu_vm_pt_first_dfs - start a deep first search + * + * @adev: amdgpu_device structure + * @vm: amdgpu_vm structure + * @start: optional cursor to start with + * @cursor: state to initialize + * + * Starts a deep first traversal of the PD/PT tree. + */ +static void amdgpu_vm_pt_first_dfs(struct amdgpu_device *adev, + struct amdgpu_vm *vm, + struct amdgpu_vm_pt_cursor *start, + struct amdgpu_vm_pt_cursor *cursor) +{ + if (start) + *cursor = *start; + else + amdgpu_vm_pt_start(adev, vm, 0, cursor); + + while (amdgpu_vm_pt_descendant(adev, cursor)) + ; +} + +/** + * amdgpu_vm_pt_continue_dfs - check if the deep first search should continue + * + * @start: starting point for the search + * @entry: current entry + * + * Returns: + * True when the search should continue, false otherwise. + */ +static bool amdgpu_vm_pt_continue_dfs(struct amdgpu_vm_pt_cursor *start, + struct amdgpu_vm_bo_base *entry) +{ + return entry && (!start || entry != start->entry); +} + +/** + * amdgpu_vm_pt_next_dfs - get the next node for a deep first search + * + * @adev: amdgpu_device structure + * @cursor: current state + * + * Move the cursor to the next node in a deep first search. + */ +static void amdgpu_vm_pt_next_dfs(struct amdgpu_device *adev, + struct amdgpu_vm_pt_cursor *cursor) +{ + if (!cursor->entry) + return; + + if (!cursor->parent) + cursor->entry = NULL; + else if (amdgpu_vm_pt_sibling(adev, cursor)) + while (amdgpu_vm_pt_descendant(adev, cursor)) + ; + else + amdgpu_vm_pt_ancestor(cursor); +} + +/* + * for_each_amdgpu_vm_pt_dfs_safe - safe deep first search of all PDs/PTs + */ +#define for_each_amdgpu_vm_pt_dfs_safe(adev, vm, start, cursor, entry) \ + for (amdgpu_vm_pt_first_dfs((adev), (vm), (start), &(cursor)), \ + (entry) = (cursor).entry, amdgpu_vm_pt_next_dfs((adev), &(cursor));\ + amdgpu_vm_pt_continue_dfs((start), (entry)); \ + (entry) = (cursor).entry, amdgpu_vm_pt_next_dfs((adev), &(cursor))) + +/** + * amdgpu_vm_pt_clear - initially clear the PDs/PTs + * + * @adev: amdgpu_device pointer + * @vm: VM to clear BO from + * @vmbo: BO to clear + * @immediate: use an immediate update + * + * Root PD needs to be reserved when calling this. + * + * Returns: + * 0 on success, errno otherwise. + */ +int amdgpu_vm_pt_clear(struct amdgpu_device *adev, struct amdgpu_vm *vm, + struct amdgpu_bo_vm *vmbo, bool immediate) +{ + unsigned int level = adev->vm_manager.root_level; + struct ttm_operation_ctx ctx = { true, false }; + struct amdgpu_vm_update_params params; + struct amdgpu_bo *ancestor = &vmbo->bo; + unsigned int entries, ats_entries; + struct amdgpu_bo *bo = &vmbo->bo; + uint64_t addr; + int r, idx; + + /* Figure out our place in the hierarchy */ + if (ancestor->parent) { + ++level; + while (ancestor->parent->parent) { + ++level; + ancestor = ancestor->parent; + } + } + + entries = amdgpu_bo_size(bo) / 8; + if (!vm->pte_support_ats) { + ats_entries = 0; + + } else if (!bo->parent) { + ats_entries = amdgpu_vm_pt_num_ats_entries(adev); + ats_entries = min(ats_entries, entries); + entries -= ats_entries; + + } else { + struct amdgpu_vm_bo_base *pt; + + pt = ancestor->vm_bo; + ats_entries = amdgpu_vm_pt_num_ats_entries(adev); + if ((pt - to_amdgpu_bo_vm(vm->root.bo)->entries) >= + ats_entries) { + ats_entries = 0; + } else { + ats_entries = entries; + entries = 0; + } + } + + r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); + if (r) + return r; + + if (vmbo->shadow) { + struct amdgpu_bo *shadow = vmbo->shadow; + + r = ttm_bo_validate(&shadow->tbo, &shadow->placement, &ctx); + if (r) + return r; + } + + if (!drm_dev_enter(adev_to_drm(adev), &idx)) + return -ENODEV; + + r = vm->update_funcs->map_table(vmbo); + if (r) + goto exit; + + memset(¶ms, 0, sizeof(params)); + params.adev = adev; + params.vm = vm; + params.immediate = immediate; + + r = vm->update_funcs->prepare(¶ms, NULL, AMDGPU_SYNC_EXPLICIT); + if (r) + goto exit; + + addr = 0; + if (ats_entries) { + uint64_t value = 0, flags; + + flags = AMDGPU_PTE_DEFAULT_ATC; + if (level != AMDGPU_VM_PTB) { + /* Handle leaf PDEs as PTEs */ + flags |= AMDGPU_PDE_PTE; + amdgpu_gmc_get_vm_pde(adev, level, &value, &flags); + } + + r = vm->update_funcs->update(¶ms, vmbo, addr, 0, + ats_entries, value, flags); + if (r) + goto exit; + + addr += ats_entries * 8; + } + + if (entries) { + uint64_t value = 0, flags = 0; + + if (adev->asic_type >= CHIP_VEGA10) { + if (level != AMDGPU_VM_PTB) { + /* Handle leaf PDEs as PTEs */ + flags |= AMDGPU_PDE_PTE; + amdgpu_gmc_get_vm_pde(adev, level, + &value, &flags); + } else { + /* Workaround for fault priority problem on GMC9 */ + flags = AMDGPU_PTE_EXECUTABLE; + } + } + + r = vm->update_funcs->update(¶ms, vmbo, addr, 0, entries, + value, flags); + if (r) + goto exit; + } + + r = vm->update_funcs->commit(¶ms, NULL); +exit: + drm_dev_exit(idx); + return r; +} + +/** + * amdgpu_vm_pt_create - create bo for PD/PT + * + * @adev: amdgpu_device pointer + * @vm: requesting vm + * @level: the page table level + * @immediate: use a immediate update + * @vmbo: pointer to the buffer object pointer + */ +int amdgpu_vm_pt_create(struct amdgpu_device *adev, struct amdgpu_vm *vm, + int level, bool immediate, struct amdgpu_bo_vm **vmbo) +{ + struct amdgpu_bo_param bp; + struct amdgpu_bo *bo; + struct dma_resv *resv; + unsigned int num_entries; + int r; + + memset(&bp, 0, sizeof(bp)); + + bp.size = amdgpu_vm_pt_size(adev, level); + bp.byte_align = AMDGPU_GPU_PAGE_SIZE; + bp.domain = AMDGPU_GEM_DOMAIN_VRAM; + bp.domain = amdgpu_bo_get_preferred_domain(adev, bp.domain); + bp.flags = AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS | + AMDGPU_GEM_CREATE_CPU_GTT_USWC; + + if (level < AMDGPU_VM_PTB) + num_entries = amdgpu_vm_pt_num_entries(adev, level); + else + num_entries = 0; + + bp.bo_ptr_size = struct_size((*vmbo), entries, num_entries); + + if (vm->use_cpu_for_update) + bp.flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED; + + bp.type = ttm_bo_type_kernel; + bp.no_wait_gpu = immediate; + if (vm->root.bo) + bp.resv = vm->root.bo->tbo.base.resv; + + r = amdgpu_bo_create_vm(adev, &bp, vmbo); + if (r) + return r; + + bo = &(*vmbo)->bo; + if (vm->is_compute_context || (adev->flags & AMD_IS_APU)) { + (*vmbo)->shadow = NULL; + return 0; + } + + if (!bp.resv) + WARN_ON(dma_resv_lock(bo->tbo.base.resv, + NULL)); + resv = bp.resv; + memset(&bp, 0, sizeof(bp)); + bp.size = amdgpu_vm_pt_size(adev, level); + bp.domain = AMDGPU_GEM_DOMAIN_GTT; + bp.flags = AMDGPU_GEM_CREATE_CPU_GTT_USWC; + bp.type = ttm_bo_type_kernel; + bp.resv = bo->tbo.base.resv; + bp.bo_ptr_size = sizeof(struct amdgpu_bo); + + r = amdgpu_bo_create(adev, &bp, &(*vmbo)->shadow); + + if (!resv) + dma_resv_unlock(bo->tbo.base.resv); + + if (r) { + amdgpu_bo_unref(&bo); + return r; + } + + (*vmbo)->shadow->parent = amdgpu_bo_ref(bo); + amdgpu_bo_add_to_shadow_list(*vmbo); + + return 0; +} + +/** + * amdgpu_vm_pt_alloc - Allocate a specific page table + * + * @adev: amdgpu_device pointer + * @vm: VM to allocate page tables for + * @cursor: Which page table to allocate + * @immediate: use an immediate update + * + * Make sure a specific page table or directory is allocated. + * + * Returns: + * 1 if page table needed to be allocated, 0 if page table was already + * allocated, negative errno if an error occurred. + */ +static int amdgpu_vm_pt_alloc(struct amdgpu_device *adev, + struct amdgpu_vm *vm, + struct amdgpu_vm_pt_cursor *cursor, + bool immediate) +{ + struct amdgpu_vm_bo_base *entry = cursor->entry; + struct amdgpu_bo *pt_bo; + struct amdgpu_bo_vm *pt; + int r; + + if (entry->bo) + return 0; + + r = amdgpu_vm_pt_create(adev, vm, cursor->level, immediate, &pt); + if (r) + return r; + + /* Keep a reference to the root directory to avoid + * freeing them up in the wrong order. + */ + pt_bo = &pt->bo; + pt_bo->parent = amdgpu_bo_ref(cursor->parent->bo); + amdgpu_vm_bo_base_init(entry, vm, pt_bo); + r = amdgpu_vm_pt_clear(adev, vm, pt, immediate); + if (r) + goto error_free_pt; + + return 0; + +error_free_pt: + amdgpu_bo_unref(&pt->shadow); + amdgpu_bo_unref(&pt_bo); + return r; +} + +/** + * amdgpu_vm_free_table - fre one PD/PT + * + * @entry: PDE to free + */ +static void amdgpu_vm_pt_free(struct amdgpu_vm_bo_base *entry) +{ + struct amdgpu_bo *shadow; + + if (!entry->bo) + return; + shadow = amdgpu_bo_shadowed(entry->bo); + entry->bo->vm_bo = NULL; + list_del(&entry->vm_status); + amdgpu_bo_unref(&shadow); + amdgpu_bo_unref(&entry->bo); +} + +/** + * amdgpu_vm_pt_free_dfs - free PD/PT levels + * + * @adev: amdgpu device structure + * @vm: amdgpu vm structure + * @start: optional cursor where to start freeing PDs/PTs + * + * Free the page directory or page table level and all sub levels. + */ +static void amdgpu_vm_pt_free_dfs(struct amdgpu_device *adev, + struct amdgpu_vm *vm, + struct amdgpu_vm_pt_cursor *start) +{ + struct amdgpu_vm_pt_cursor cursor; + struct amdgpu_vm_bo_base *entry; + + vm->bulk_moveable = false; + + for_each_amdgpu_vm_pt_dfs_safe(adev, vm, start, cursor, entry) + amdgpu_vm_pt_free(entry); + + if (start) + amdgpu_vm_pt_free(start->entry); +} + +/** + * amdgpu_vm_pt_free_root - free root PD + * @adev: amdgpu device structure + * @vm: amdgpu vm structure + * + * Free the root page directory and everything below it. + */ +void amdgpu_vm_pt_free_root(struct amdgpu_device *adev, struct amdgpu_vm *vm) +{ + amdgpu_vm_pt_free_dfs(adev, vm, NULL); +} + +/** + * amdgpu_vm_pt_is_root_clean - check if a root PD is clean + * + * @adev: amdgpu_device pointer + * @vm: the VM to check + * + * Check all entries of the root PD, if any subsequent PDs are allocated, + * it means there are page table creating and filling, and is no a clean + * VM + * + * Returns: + * 0 if this VM is clean + */ +bool amdgpu_vm_pt_is_root_clean(struct amdgpu_device *adev, + struct amdgpu_vm *vm) +{ + enum amdgpu_vm_level root = adev->vm_manager.root_level; + unsigned int entries = amdgpu_vm_pt_num_entries(adev, root); + unsigned int i = 0; + + for (i = 0; i < entries; i++) { + if (to_amdgpu_bo_vm(vm->root.bo)->entries[i].bo) + return false; + } + return true; +} + +/** + * amdgpu_vm_pde_update - update a single level in the hierarchy + * + * @params: parameters for the update + * @entry: entry to update + * + * Makes sure the requested entry in parent is up to date. + */ +int amdgpu_vm_pde_update(struct amdgpu_vm_update_params *params, + struct amdgpu_vm_bo_base *entry) +{ + struct amdgpu_vm_bo_base *parent = amdgpu_vm_pt_parent(entry); + struct amdgpu_bo *bo = parent->bo, *pbo; + struct amdgpu_vm *vm = params->vm; + uint64_t pde, pt, flags; + unsigned int level; + + for (level = 0, pbo = bo->parent; pbo; ++level) + pbo = pbo->parent; + + level += params->adev->vm_manager.root_level; + amdgpu_gmc_get_pde_for_bo(entry->bo, level, &pt, &flags); + pde = (entry - to_amdgpu_bo_vm(parent->bo)->entries) * 8; + return vm->update_funcs->update(params, to_amdgpu_bo_vm(bo), pde, pt, + 1, 0, flags); +} + +/* + * amdgpu_vm_pte_update_flags - figure out flags for PTE updates + * + * Make sure to set the right flags for the PTEs at the desired level. + */ +static void amdgpu_vm_pte_update_flags(struct amdgpu_vm_update_params *params, + struct amdgpu_bo_vm *pt, + unsigned int level, + uint64_t pe, uint64_t addr, + unsigned int count, uint32_t incr, + uint64_t flags) + +{ + if (level != AMDGPU_VM_PTB) { + flags |= AMDGPU_PDE_PTE; + amdgpu_gmc_get_vm_pde(params->adev, level, &addr, &flags); + + } else if (params->adev->asic_type >= CHIP_VEGA10 && + !(flags & AMDGPU_PTE_VALID) && + !(flags & AMDGPU_PTE_PRT)) { + + /* Workaround for fault priority problem on GMC9 */ + flags |= AMDGPU_PTE_EXECUTABLE; + } + + params->vm->update_funcs->update(params, pt, pe, addr, count, incr, + flags); +} + +/** + * amdgpu_vm_pte_fragment - get fragment for PTEs + * + * @params: see amdgpu_vm_update_params definition + * @start: first PTE to handle + * @end: last PTE to handle + * @flags: hw mapping flags + * @frag: resulting fragment size + * @frag_end: end of this fragment + * + * Returns the first possible fragment for the start and end address. + */ +static void amdgpu_vm_pte_fragment(struct amdgpu_vm_update_params *params, + uint64_t start, uint64_t end, uint64_t flags, + unsigned int *frag, uint64_t *frag_end) +{ + /** + * The MC L1 TLB supports variable sized pages, based on a fragment + * field in the PTE. When this field is set to a non-zero value, page + * granularity is increased from 4KB to (1 << (12 + frag)). The PTE + * flags are considered valid for all PTEs within the fragment range + * and corresponding mappings are assumed to be physically contiguous. + * + * The L1 TLB can store a single PTE for the whole fragment, + * significantly increasing the space available for translation + * caching. This leads to large improvements in throughput when the + * TLB is under pressure. + * + * The L2 TLB distributes small and large fragments into two + * asymmetric partitions. The large fragment cache is significantly + * larger. Thus, we try to use large fragments wherever possible. + * Userspace can support this by aligning virtual base address and + * allocation size to the fragment size. + * + * Starting with Vega10 the fragment size only controls the L1. The L2 + * is now directly feed with small/huge/giant pages from the walker. + */ + unsigned int max_frag; + + if (params->adev->asic_type < CHIP_VEGA10) + max_frag = params->adev->vm_manager.fragment_size; + else + max_frag = 31; + + /* system pages are non continuously */ + if (params->pages_addr) { + *frag = 0; + *frag_end = end; + return; + } + + /* This intentionally wraps around if no bit is set */ + *frag = min_t(unsigned int, ffs(start) - 1, fls64(end - start) - 1); + if (*frag >= max_frag) { + *frag = max_frag; + *frag_end = end & ~((1ULL << max_frag) - 1); + } else { + *frag_end = start + (1 << *frag); + } +} + +/** + * amdgpu_vm_ptes_update - make sure that page tables are valid + * + * @params: see amdgpu_vm_update_params definition + * @start: start of GPU address range + * @end: end of GPU address range + * @dst: destination address to map to, the next dst inside the function + * @flags: mapping flags + * + * Update the page tables in the range @start - @end. + * + * Returns: + * 0 for success, -EINVAL for failure. + */ +int amdgpu_vm_ptes_update(struct amdgpu_vm_update_params *params, + uint64_t start, uint64_t end, + uint64_t dst, uint64_t flags) +{ + struct amdgpu_device *adev = params->adev; + struct amdgpu_vm_pt_cursor cursor; + uint64_t frag_start = start, frag_end; + unsigned int frag; + int r; + + /* figure out the initial fragment */ + amdgpu_vm_pte_fragment(params, frag_start, end, flags, &frag, + &frag_end); + + /* walk over the address space and update the PTs */ + amdgpu_vm_pt_start(adev, params->vm, start, &cursor); + while (cursor.pfn < end) { + unsigned int shift, parent_shift, mask; + uint64_t incr, entry_end, pe_start; + struct amdgpu_bo *pt; + + if (!params->unlocked) { + /* make sure that the page tables covering the + * address range are actually allocated + */ + r = amdgpu_vm_pt_alloc(params->adev, params->vm, + &cursor, params->immediate); + if (r) + return r; + } + + shift = amdgpu_vm_pt_level_shift(adev, cursor.level); + parent_shift = amdgpu_vm_pt_level_shift(adev, cursor.level - 1); + if (params->unlocked) { + /* Unlocked updates are only allowed on the leaves */ + if (amdgpu_vm_pt_descendant(adev, &cursor)) + continue; + } else if (adev->asic_type < CHIP_VEGA10 && + (flags & AMDGPU_PTE_VALID)) { + /* No huge page support before GMC v9 */ + if (cursor.level != AMDGPU_VM_PTB) { + if (!amdgpu_vm_pt_descendant(adev, &cursor)) + return -ENOENT; + continue; + } + } else if (frag < shift) { + /* We can't use this level when the fragment size is + * smaller than the address shift. Go to the next + * child entry and try again. + */ + if (amdgpu_vm_pt_descendant(adev, &cursor)) + continue; + } else if (frag >= parent_shift) { + /* If the fragment size is even larger than the parent + * shift we should go up one level and check it again. + */ + if (!amdgpu_vm_pt_ancestor(&cursor)) + return -EINVAL; + continue; + } + + pt = cursor.entry->bo; + if (!pt) { + /* We need all PDs and PTs for mapping something, */ + if (flags & AMDGPU_PTE_VALID) + return -ENOENT; + + /* but unmapping something can happen at a higher + * level. + */ + if (!amdgpu_vm_pt_ancestor(&cursor)) + return -EINVAL; + + pt = cursor.entry->bo; + shift = parent_shift; + frag_end = max(frag_end, ALIGN(frag_start + 1, + 1ULL << shift)); + } + + /* Looks good so far, calculate parameters for the update */ + incr = (uint64_t)AMDGPU_GPU_PAGE_SIZE << shift; + mask = amdgpu_vm_pt_entries_mask(adev, cursor.level); + pe_start = ((cursor.pfn >> shift) & mask) * 8; + entry_end = ((uint64_t)mask + 1) << shift; + entry_end += cursor.pfn & ~(entry_end - 1); + entry_end = min(entry_end, end); + + do { + struct amdgpu_vm *vm = params->vm; + uint64_t upd_end = min(entry_end, frag_end); + unsigned int nptes = (upd_end - frag_start) >> shift; + uint64_t upd_flags = flags | AMDGPU_PTE_FRAG(frag); + + /* This can happen when we set higher level PDs to + * silent to stop fault floods. + */ + nptes = max(nptes, 1u); + + trace_amdgpu_vm_update_ptes(params, frag_start, upd_end, + min(nptes, 32u), dst, incr, + upd_flags, + vm->task_info.pid, + vm->immediate.fence_context); + amdgpu_vm_pte_update_flags(params, to_amdgpu_bo_vm(pt), + cursor.level, pe_start, dst, + nptes, incr, upd_flags); + + pe_start += nptes * 8; + dst += nptes * incr; + + frag_start = upd_end; + if (frag_start >= frag_end) { + /* figure out the next fragment */ + amdgpu_vm_pte_fragment(params, frag_start, end, + flags, &frag, &frag_end); + if (frag < shift) + break; + } + } while (frag_start < entry_end); + + if (amdgpu_vm_pt_descendant(adev, &cursor)) { + /* Free all child entries. + * Update the tables with the flags and addresses and free up subsequent + * tables in the case of huge pages or freed up areas. + * This is the maximum you can free, because all other page tables are not + * completely covered by the range and so potentially still in use. + */ + while (cursor.pfn < frag_start) { + /* Make sure previous mapping is freed */ + if (cursor.entry->bo) { + params->table_freed = true; + amdgpu_vm_pt_free_dfs(adev, params->vm, + &cursor); + } + amdgpu_vm_pt_next(adev, &cursor); + } + + } else if (frag >= shift) { + /* or just move on to the next on the same level. */ + amdgpu_vm_pt_next(adev, &cursor); + } + } + + return 0; +} -- cgit v1.2.3 From e997b82745a5b2419bc5b72735811d7162ac994a Mon Sep 17 00:00:00 2001 From: Christian König Date: Tue, 8 Mar 2022 15:06:49 +0100 Subject: drm/amdgpu: simplify VM update tracking a bit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Store the 64bit sequence directly. Makes it simpler to use and saves a bit of fence reference counting overhead. Signed-off-by: Christian König Acked-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c | 40 +++++++++----------------------- drivers/gpu/drm/amd/amdgpu/amdgpu_ids.h | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c | 9 +++---- drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h | 2 +- 4 files changed, 16 insertions(+), 37 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c index 92a70fb57fa3..4f1399573107 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c @@ -276,19 +276,15 @@ static int amdgpu_vmid_grab_reserved(struct amdgpu_vm *vm, struct amdgpu_device *adev = ring->adev; unsigned vmhub = ring->funcs->vmhub; uint64_t fence_context = adev->fence_context + ring->idx; - struct dma_fence *updates = sync->last_vm_update; bool needs_flush = vm->use_cpu_for_update; - int r = 0; + uint64_t updates = sync->last_vm_update; + int r; *id = vm->reserved_vmid[vmhub]; - if (updates && (*id)->flushed_updates && - updates->context == (*id)->flushed_updates->context && - !dma_fence_is_later(updates, (*id)->flushed_updates)) - updates = NULL; - if ((*id)->owner != vm->immediate.fence_context || - job->vm_pd_addr != (*id)->pd_gpu_addr || - updates || !(*id)->last_flush || + (*id)->pd_gpu_addr != job->vm_pd_addr || + (*id)->flushed_updates < updates || + !(*id)->last_flush || ((*id)->last_flush->context != fence_context && !dma_fence_is_signaled((*id)->last_flush))) { struct dma_fence *tmp; @@ -302,8 +298,7 @@ static int amdgpu_vmid_grab_reserved(struct amdgpu_vm *vm, tmp = amdgpu_sync_peek_fence(&(*id)->active, ring); if (tmp) { *id = NULL; - r = amdgpu_sync_fence(sync, tmp); - return r; + return amdgpu_sync_fence(sync, tmp); } needs_flush = true; } @@ -315,10 +310,7 @@ static int amdgpu_vmid_grab_reserved(struct amdgpu_vm *vm, if (r) return r; - if (updates) { - dma_fence_put((*id)->flushed_updates); - (*id)->flushed_updates = dma_fence_get(updates); - } + (*id)->flushed_updates = updates; job->vm_needs_flush = needs_flush; return 0; } @@ -346,7 +338,7 @@ static int amdgpu_vmid_grab_used(struct amdgpu_vm *vm, unsigned vmhub = ring->funcs->vmhub; struct amdgpu_vmid_mgr *id_mgr = &adev->vm_manager.id_mgr[vmhub]; uint64_t fence_context = adev->fence_context + ring->idx; - struct dma_fence *updates = sync->last_vm_update; + uint64_t updates = sync->last_vm_update; int r; job->vm_needs_flush = vm->use_cpu_for_update; @@ -354,7 +346,6 @@ static int amdgpu_vmid_grab_used(struct amdgpu_vm *vm, /* Check if we can use a VMID already assigned to this VM */ list_for_each_entry_reverse((*id), &id_mgr->ids_lru, list) { bool needs_flush = vm->use_cpu_for_update; - struct dma_fence *flushed; /* Check all the prerequisites to using this VMID */ if ((*id)->owner != vm->immediate.fence_context) @@ -368,8 +359,7 @@ static int amdgpu_vmid_grab_used(struct amdgpu_vm *vm, !dma_fence_is_signaled((*id)->last_flush))) needs_flush = true; - flushed = (*id)->flushed_updates; - if (updates && (!flushed || dma_fence_is_later(updates, flushed))) + if ((*id)->flushed_updates < updates) needs_flush = true; if (needs_flush && !adev->vm_manager.concurrent_flush) @@ -382,11 +372,7 @@ static int amdgpu_vmid_grab_used(struct amdgpu_vm *vm, if (r) return r; - if (updates && (!flushed || dma_fence_is_later(updates, flushed))) { - dma_fence_put((*id)->flushed_updates); - (*id)->flushed_updates = dma_fence_get(updates); - } - + (*id)->flushed_updates = updates; job->vm_needs_flush |= needs_flush; return 0; } @@ -432,8 +418,6 @@ int amdgpu_vmid_grab(struct amdgpu_vm *vm, struct amdgpu_ring *ring, goto error; if (!id) { - struct dma_fence *updates = sync->last_vm_update; - /* Still no ID to use? Then use the idle one found earlier */ id = idle; @@ -442,8 +426,7 @@ int amdgpu_vmid_grab(struct amdgpu_vm *vm, struct amdgpu_ring *ring, if (r) goto error; - dma_fence_put(id->flushed_updates); - id->flushed_updates = dma_fence_get(updates); + id->flushed_updates = sync->last_vm_update; job->vm_needs_flush = true; } @@ -610,7 +593,6 @@ void amdgpu_vmid_mgr_fini(struct amdgpu_device *adev) struct amdgpu_vmid *id = &id_mgr->ids[j]; amdgpu_sync_free(&id->active); - dma_fence_put(id->flushed_updates); dma_fence_put(id->last_flush); dma_fence_put(id->pasid_mapping); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.h index 0c3b4fa1f936..06c8a0034fa5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.h @@ -47,7 +47,7 @@ struct amdgpu_vmid { uint64_t pd_gpu_addr; /* last flushed PD/PT update */ - struct dma_fence *flushed_updates; + uint64_t flushed_updates; uint32_t current_gpu_reset_count; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c index 40e06745fae9..05e1af9998a3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c @@ -51,7 +51,7 @@ static struct kmem_cache *amdgpu_sync_slab; void amdgpu_sync_create(struct amdgpu_sync *sync) { hash_init(sync->fences); - sync->last_vm_update = NULL; + sync->last_vm_update = 0; } /** @@ -184,7 +184,7 @@ int amdgpu_sync_vm_fence(struct amdgpu_sync *sync, struct dma_fence *fence) if (!fence) return 0; - amdgpu_sync_keep_later(&sync->last_vm_update, fence); + sync->last_vm_update = max(sync->last_vm_update, fence->seqno); return amdgpu_sync_fence(sync, fence); } @@ -376,8 +376,7 @@ int amdgpu_sync_clone(struct amdgpu_sync *source, struct amdgpu_sync *clone) } } - dma_fence_put(clone->last_vm_update); - clone->last_vm_update = dma_fence_get(source->last_vm_update); + clone->last_vm_update = source->last_vm_update; return 0; } @@ -419,8 +418,6 @@ void amdgpu_sync_free(struct amdgpu_sync *sync) dma_fence_put(e->fence); kmem_cache_free(amdgpu_sync_slab, e); } - - dma_fence_put(sync->last_vm_update); } /** diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h index 7c0fe20c470d..876c1ee8869c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h @@ -43,7 +43,7 @@ enum amdgpu_sync_mode { */ struct amdgpu_sync { DECLARE_HASHTABLE(fences, 4); - struct dma_fence *last_vm_update; + uint64_t last_vm_update; }; void amdgpu_sync_create(struct amdgpu_sync *sync); -- cgit v1.2.3 From 5255e146c99a677d4d55fdb988544bd20c539a0b Mon Sep 17 00:00:00 2001 From: Christian König Date: Tue, 15 Mar 2022 15:27:45 +0100 Subject: drm/amdgpu: rework TLB flushing MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Instead of tracking the VM updates through the dependencies just use a sequence counter for page table updates which indicates the need to flush the TLB. This reduces the need to flush the TLB drastically. v2: squash in NULL check fix (Christian) Signed-off-by: Christian König Acked-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 8 ++--- drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c | 6 ++-- drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c | 20 ----------- drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h | 2 -- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 57 ++++++++++++++++++++++++++++++-- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 15 +++++++++ 6 files changed, 76 insertions(+), 32 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 970b065e9a6b..88681abc5f8c 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -810,7 +810,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (r) return r; - r = amdgpu_sync_vm_fence(&p->job->sync, fpriv->prt_va->last_pt_update); + r = amdgpu_sync_fence(&p->job->sync, fpriv->prt_va->last_pt_update); if (r) return r; @@ -821,7 +821,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (r) return r; - r = amdgpu_sync_vm_fence(&p->job->sync, bo_va->last_pt_update); + r = amdgpu_sync_fence(&p->job->sync, bo_va->last_pt_update); if (r) return r; } @@ -840,7 +840,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (r) return r; - r = amdgpu_sync_vm_fence(&p->job->sync, bo_va->last_pt_update); + r = amdgpu_sync_fence(&p->job->sync, bo_va->last_pt_update); if (r) return r; } @@ -853,7 +853,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (r) return r; - r = amdgpu_sync_vm_fence(&p->job->sync, vm->last_update); + r = amdgpu_sync_fence(&p->job->sync, vm->last_update); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c index 4f1399573107..558f90e11d78 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c @@ -277,7 +277,7 @@ static int amdgpu_vmid_grab_reserved(struct amdgpu_vm *vm, unsigned vmhub = ring->funcs->vmhub; uint64_t fence_context = adev->fence_context + ring->idx; bool needs_flush = vm->use_cpu_for_update; - uint64_t updates = sync->last_vm_update; + uint64_t updates = amdgpu_vm_tlb_seq(vm); int r; *id = vm->reserved_vmid[vmhub]; @@ -338,7 +338,7 @@ static int amdgpu_vmid_grab_used(struct amdgpu_vm *vm, unsigned vmhub = ring->funcs->vmhub; struct amdgpu_vmid_mgr *id_mgr = &adev->vm_manager.id_mgr[vmhub]; uint64_t fence_context = adev->fence_context + ring->idx; - uint64_t updates = sync->last_vm_update; + uint64_t updates = amdgpu_vm_tlb_seq(vm); int r; job->vm_needs_flush = vm->use_cpu_for_update; @@ -426,7 +426,7 @@ int amdgpu_vmid_grab(struct amdgpu_vm *vm, struct amdgpu_ring *ring, if (r) goto error; - id->flushed_updates = sync->last_vm_update; + id->flushed_updates = amdgpu_vm_tlb_seq(vm); job->vm_needs_flush = true; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c index 05e1af9998a3..edd49dd27422 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c @@ -51,7 +51,6 @@ static struct kmem_cache *amdgpu_sync_slab; void amdgpu_sync_create(struct amdgpu_sync *sync) { hash_init(sync->fences); - sync->last_vm_update = 0; } /** @@ -171,23 +170,6 @@ int amdgpu_sync_fence(struct amdgpu_sync *sync, struct dma_fence *f) return 0; } -/** - * amdgpu_sync_vm_fence - remember to sync to this VM fence - * - * @sync: sync object to add fence to - * @fence: the VM fence to add - * - * Add the fence to the sync object and remember it as VM update. - */ -int amdgpu_sync_vm_fence(struct amdgpu_sync *sync, struct dma_fence *fence) -{ - if (!fence) - return 0; - - sync->last_vm_update = max(sync->last_vm_update, fence->seqno); - return amdgpu_sync_fence(sync, fence); -} - /* Determine based on the owner and mode if we should sync to a fence or not */ static bool amdgpu_sync_test_fence(struct amdgpu_device *adev, enum amdgpu_sync_mode mode, @@ -376,8 +358,6 @@ int amdgpu_sync_clone(struct amdgpu_sync *source, struct amdgpu_sync *clone) } } - clone->last_vm_update = source->last_vm_update; - return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h index 876c1ee8869c..2d5c613cda10 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.h @@ -43,12 +43,10 @@ enum amdgpu_sync_mode { */ struct amdgpu_sync { DECLARE_HASHTABLE(fences, 4); - uint64_t last_vm_update; }; void amdgpu_sync_create(struct amdgpu_sync *sync); int amdgpu_sync_fence(struct amdgpu_sync *sync, struct dma_fence *f); -int amdgpu_sync_vm_fence(struct amdgpu_sync *sync, struct dma_fence *fence); int amdgpu_sync_resv(struct amdgpu_device *adev, struct amdgpu_sync *sync, struct dma_resv *resv, enum amdgpu_sync_mode mode, void *owner); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index accebbac2f29..ecdcf652ba41 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -88,6 +88,21 @@ struct amdgpu_prt_cb { struct dma_fence_cb cb; }; +/** + * amdgpu_vm_tlb_seq_cb - Helper to increment the TLB flush sequence + */ +struct amdgpu_vm_tlb_seq_cb { + /** + * @vm: pointer to the amdgpu_vm structure to set the fence sequence on + */ + struct amdgpu_vm *vm; + + /** + * @cb: callback + */ + struct dma_fence_cb cb; +}; + /** * amdgpu_vm_set_pasid - manage pasid and vm ptr mapping * @@ -760,6 +775,23 @@ error: return r; } +/** + * amdgpu_vm_tlb_seq_cb - make sure to increment tlb sequence + * @fence: unused + * @cb: the callback structure + * + * Increments the tlb sequence to make sure that future CS execute a VM flush. + */ +static void amdgpu_vm_tlb_seq_cb(struct dma_fence *fence, + struct dma_fence_cb *cb) +{ + struct amdgpu_vm_tlb_seq_cb *tlb_cb; + + tlb_cb = container_of(cb, typeof(*tlb_cb), cb); + atomic64_inc(&tlb_cb->vm->tlb_seq); + kfree(tlb_cb); +} + /** * amdgpu_vm_bo_update_mapping - update a mapping in the vm page table * @@ -795,6 +827,7 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, bool *table_freed) { struct amdgpu_vm_update_params params; + struct amdgpu_vm_tlb_seq_cb *tlb_cb; struct amdgpu_res_cursor cursor; enum amdgpu_sync_mode sync_mode; int r, idx; @@ -802,6 +835,12 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, if (!drm_dev_enter(adev_to_drm(adev), &idx)) return -ENODEV; + tlb_cb = kmalloc(sizeof(*tlb_cb), GFP_KERNEL); + if (!tlb_cb) { + r = -ENOMEM; + goto error_unlock; + } + memset(¶ms, 0, sizeof(params)); params.adev = adev; params.vm = vm; @@ -820,7 +859,7 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, amdgpu_vm_eviction_lock(vm); if (vm->evicting) { r = -EBUSY; - goto error_unlock; + goto error_free; } if (!unlocked && !dma_fence_is_signaled(vm->last_unlocked)) { @@ -833,7 +872,7 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, r = vm->update_funcs->prepare(¶ms, resv, sync_mode); if (r) - goto error_unlock; + goto error_free; amdgpu_res_first(pages_addr ? NULL : res, offset, (last - start + 1) * AMDGPU_GPU_PAGE_SIZE, &cursor); @@ -882,7 +921,7 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, tmp = start + num_entries; r = amdgpu_vm_ptes_update(¶ms, start, tmp, addr, flags); if (r) - goto error_unlock; + goto error_free; amdgpu_res_next(&cursor, num_entries * AMDGPU_GPU_PAGE_SIZE); start = tmp; @@ -890,9 +929,21 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, r = vm->update_funcs->commit(¶ms, fence); + if (!unlocked && (!(flags & AMDGPU_PTE_VALID) || params.table_freed)) { + tlb_cb->vm = vm; + if (!fence || !*fence || + dma_fence_add_callback(*fence, &tlb_cb->cb, + amdgpu_vm_tlb_seq_cb)) + amdgpu_vm_tlb_seq_cb(NULL, &tlb_cb->cb); + tlb_cb = NULL; + } + if (table_freed) *table_freed = *table_freed || params.table_freed; +error_free: + kfree(tlb_cb); + error_unlock: amdgpu_vm_eviction_unlock(vm); drm_dev_exit(idx); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index 1731681914f5..38a1eab1ff74 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -284,6 +284,9 @@ struct amdgpu_vm { struct drm_sched_entity immediate; struct drm_sched_entity delayed; + /* Last finished delayed update */ + atomic64_t tlb_seq; + /* Last unlocked submission to the scheduler entities */ struct dma_fence *last_unlocked; @@ -478,4 +481,16 @@ int amdgpu_vm_ptes_update(struct amdgpu_vm_update_params *params, void amdgpu_debugfs_vm_bo_info(struct amdgpu_vm *vm, struct seq_file *m); #endif +/** + * amdgpu_vm_tlb_seq - return tlb flush sequence number + * @vm: the amdgpu_vm structure to query + * + * Returns the tlb flush sequence number which indicates that the VM TLBs needs + * to be invalidated whenever the sequence number change. + */ +static inline uint64_t amdgpu_vm_tlb_seq(struct amdgpu_vm *vm) +{ + return atomic64_read(&vm->tlb_seq); +} + #endif -- cgit v1.2.3 From bffa91dadf599155200b3efb46217ec108078a30 Mon Sep 17 00:00:00 2001 From: Christian König Date: Thu, 17 Mar 2022 09:53:55 +0100 Subject: drm/amdkfd: start using tlb_seq from the VM subsystem MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Instead of trying to figure out if a TLB flush is necessary or not use the information provided by the VM subsystem now. Signed-off-by: Christian König Reviewed-by: Philip Yang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_priv.h | 1 + drivers/gpu/drm/amd/amdkfd/kfd_process.c | 7 +++++++ 2 files changed, 8 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index f36062be9ca8..945982a5d688 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -705,6 +705,7 @@ struct kfd_process_device { /* VM context for GPUVM allocations */ struct file *drm_file; void *drm_priv; + uint64_t tlb_seq; /* GPUVM allocations storage */ struct idr alloc_idr; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index 59c04b2d383b..4a8a047b7593 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -1560,6 +1560,7 @@ int kfd_process_device_init_vm(struct kfd_process_device *pdd, return ret; } pdd->drm_priv = drm_file->private_data; + pdd->tlb_seq = 0; ret = kfd_process_device_reserve_ib_mem(pdd); if (ret) @@ -1949,8 +1950,14 @@ int kfd_reserved_mem_mmap(struct kfd_dev *dev, struct kfd_process *process, void kfd_flush_tlb(struct kfd_process_device *pdd, enum TLB_FLUSH_TYPE type) { + struct amdgpu_vm *vm = drm_priv_to_vm(pdd->drm_priv); + uint64_t tlb_seq = amdgpu_vm_tlb_seq(vm); struct kfd_dev *dev = pdd->dev; + if (pdd->tlb_seq == tlb_seq) + return; + + pdd->tlb_seq = tlb_seq; if (dev->dqm->sched_policy == KFD_SCHED_POLICY_NO_HWS) { /* Nothing to flush until a VMID is assigned, which * only happens when the first queue is created. -- cgit v1.2.3 From 4d30a83c740e9904c7f54f071ec121f9e6932f63 Mon Sep 17 00:00:00 2001 From: Christian König Date: Thu, 17 Mar 2022 14:08:36 +0100 Subject: drm/amdkfd: use tlb_seq from the VM subsystem for SVM as well v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Instead of hand rolling the table_freed parameter. v2: add some changes suggested by Philip Signed-off-by: Christian König Reviewed-by: Philip Yang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 5 ++--- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 18 ++++++++---------- drivers/gpu/drm/amd/amdkfd/kfd_chardev.c | 19 ++++++++----------- drivers/gpu/drm/amd/amdkfd/kfd_process.c | 2 +- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 7 +++---- 5 files changed, 22 insertions(+), 29 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index 0838926a8ef0..f8b9f27adcf5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -273,9 +273,8 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu( int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( struct amdgpu_device *adev, struct kgd_mem *mem, void *drm_priv, uint64_t *size); -int amdgpu_amdkfd_gpuvm_map_memory_to_gpu( - struct amdgpu_device *adev, struct kgd_mem *mem, void *drm_priv, - bool *table_freed); +int amdgpu_amdkfd_gpuvm_map_memory_to_gpu(struct amdgpu_device *adev, + struct kgd_mem *mem, void *drm_priv); int amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu( struct amdgpu_device *adev, struct kgd_mem *mem, void *drm_priv); int amdgpu_amdkfd_gpuvm_sync_memory( diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index cd89d2e46852..57b521bb1eec 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -1093,8 +1093,7 @@ static void unmap_bo_from_gpuvm(struct kgd_mem *mem, static int update_gpuvm_pte(struct kgd_mem *mem, struct kfd_mem_attachment *entry, - struct amdgpu_sync *sync, - bool *table_freed) + struct amdgpu_sync *sync) { struct amdgpu_bo_va *bo_va = entry->bo_va; struct amdgpu_device *adev = entry->adev; @@ -1105,7 +1104,7 @@ static int update_gpuvm_pte(struct kgd_mem *mem, return ret; /* Update the page tables */ - ret = amdgpu_vm_bo_update(adev, bo_va, false, table_freed); + ret = amdgpu_vm_bo_update(adev, bo_va, false, NULL); if (ret) { pr_err("amdgpu_vm_bo_update failed\n"); return ret; @@ -1117,8 +1116,7 @@ static int update_gpuvm_pte(struct kgd_mem *mem, static int map_bo_to_gpuvm(struct kgd_mem *mem, struct kfd_mem_attachment *entry, struct amdgpu_sync *sync, - bool no_update_pte, - bool *table_freed) + bool no_update_pte) { int ret; @@ -1135,7 +1133,7 @@ static int map_bo_to_gpuvm(struct kgd_mem *mem, if (no_update_pte) return 0; - ret = update_gpuvm_pte(mem, entry, sync, table_freed); + ret = update_gpuvm_pte(mem, entry, sync); if (ret) { pr_err("update_gpuvm_pte() failed\n"); goto update_gpuvm_pte_failed; @@ -1745,7 +1743,7 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu( int amdgpu_amdkfd_gpuvm_map_memory_to_gpu( struct amdgpu_device *adev, struct kgd_mem *mem, - void *drm_priv, bool *table_freed) + void *drm_priv) { struct amdgpu_vm *avm = drm_priv_to_vm(drm_priv); int ret; @@ -1832,7 +1830,7 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu( entry->va, entry->va + bo_size, entry); ret = map_bo_to_gpuvm(mem, entry, ctx.sync, - is_invalid_userptr, table_freed); + is_invalid_userptr); if (ret) { pr_err("Failed to map bo to gpuvm\n"); goto out_unreserve; @@ -2300,7 +2298,7 @@ static int validate_invalid_user_pages(struct amdkfd_process_info *process_info) continue; kfd_mem_dmaunmap_attachment(mem, attachment); - ret = update_gpuvm_pte(mem, attachment, &sync, NULL); + ret = update_gpuvm_pte(mem, attachment, &sync); if (ret) { pr_err("%s: update PTE failed\n", __func__); /* make sure this gets validated again */ @@ -2506,7 +2504,7 @@ int amdgpu_amdkfd_gpuvm_restore_process_bos(void *info, struct dma_fence **ef) continue; kfd_mem_dmaunmap_attachment(mem, attachment); - ret = update_gpuvm_pte(mem, attachment, &sync_obj, NULL); + ret = update_gpuvm_pte(mem, attachment, &sync_obj); if (ret) { pr_debug("Memory eviction: update PTE failed. Try again\n"); goto validate_map_fail; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index 607f65ab39ac..60438193c0c1 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -1146,7 +1146,6 @@ static int kfd_ioctl_map_memory_to_gpu(struct file *filep, long err = 0; int i; uint32_t *devices_arr = NULL; - bool table_freed = false; if (!args->n_devices) { pr_debug("Device IDs array empty\n"); @@ -1208,7 +1207,7 @@ static int kfd_ioctl_map_memory_to_gpu(struct file *filep, err = amdgpu_amdkfd_gpuvm_map_memory_to_gpu( peer_pdd->dev->adev, (struct kgd_mem *)mem, - peer_pdd->drm_priv, &table_freed); + peer_pdd->drm_priv); if (err) { struct pci_dev *pdev = peer_pdd->dev->adev->pdev; @@ -1233,13 +1232,11 @@ static int kfd_ioctl_map_memory_to_gpu(struct file *filep, } /* Flush TLBs after waiting for the page table updates to complete */ - if (table_freed || !kfd_flush_tlb_after_unmap(dev)) { - for (i = 0; i < args->n_devices; i++) { - peer_pdd = kfd_process_device_data_by_id(p, devices_arr[i]); - if (WARN_ON_ONCE(!peer_pdd)) - continue; - kfd_flush_tlb(peer_pdd, TLB_FLUSH_LEGACY); - } + for (i = 0; i < args->n_devices; i++) { + peer_pdd = kfd_process_device_data_by_id(p, devices_arr[i]); + if (WARN_ON_ONCE(!peer_pdd)) + continue; + kfd_flush_tlb(peer_pdd, TLB_FLUSH_LEGACY); } kfree(devices_arr); @@ -2206,8 +2203,8 @@ static int criu_restore_bo(struct kfd_process *p, if (IS_ERR(peer_pdd)) return PTR_ERR(peer_pdd); - ret = amdgpu_amdkfd_gpuvm_map_memory_to_gpu(peer->adev, kgd_mem, peer_pdd->drm_priv, - NULL); + ret = amdgpu_amdkfd_gpuvm_map_memory_to_gpu(peer->adev, kgd_mem, + peer_pdd->drm_priv); if (ret) { pr_err("Failed to map to gpu %d/%d\n", j, p->n_pdds); return ret; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index 4a8a047b7593..ac8123c1ee8f 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -722,7 +722,7 @@ static int kfd_process_alloc_gpuvm(struct kfd_process_device *pdd, goto err_alloc_mem; err = amdgpu_amdkfd_gpuvm_map_memory_to_gpu(kdev->adev, *mem, - pdd->drm_priv, NULL); + pdd->drm_priv); if (err) goto err_map_mem; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index 3b8856b4cece..bf6354f539d8 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -1243,7 +1243,6 @@ svm_range_map_to_gpu(struct kfd_process_device *pdd, struct svm_range *prange, { struct amdgpu_device *adev = pdd->dev->adev; struct amdgpu_vm *vm = drm_priv_to_vm(pdd->drm_priv); - bool table_freed = false; uint64_t pte_flags; unsigned long last_start; int last_domain; @@ -1284,7 +1283,7 @@ svm_range_map_to_gpu(struct kfd_process_device *pdd, struct svm_range *prange, last_start - prange->start, NULL, dma_addr, &vm->last_update, - &table_freed); + NULL); for (j = last_start - prange->start; j <= i; j++) dma_addr[j] |= last_domain; @@ -1306,8 +1305,6 @@ svm_range_map_to_gpu(struct kfd_process_device *pdd, struct svm_range *prange, if (fence) *fence = dma_fence_get(vm->last_update); - if (table_freed) - kfd_flush_tlb(pdd, TLB_FLUSH_LEGACY); out: return r; } @@ -1363,6 +1360,8 @@ svm_range_map_to_gpus(struct svm_range *prange, unsigned long offset, break; } } + + kfd_flush_tlb(pdd, TLB_FLUSH_LEGACY); } return r; -- cgit v1.2.3 From 8f8cc3fb43508a2b1682e3809d6d39ce1871a5ee Mon Sep 17 00:00:00 2001 From: Christian König Date: Thu, 17 Mar 2022 14:29:22 +0100 Subject: drm/amdgpu: remove table_freed param from the VM code MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Better to leave the decision when to flush the VM changes in the TLB to the VM code. Signed-off-by: Christian König Reviewed-by: Philip Yang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 6 +++--- drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 21 +++++++-------------- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 4 ++-- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 5 ++--- 6 files changed, 16 insertions(+), 24 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index 57b521bb1eec..8b14c55a0793 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -1104,7 +1104,7 @@ static int update_gpuvm_pte(struct kgd_mem *mem, return ret; /* Update the page tables */ - ret = amdgpu_vm_bo_update(adev, bo_va, false, NULL); + ret = amdgpu_vm_bo_update(adev, bo_va, false); if (ret) { pr_err("amdgpu_vm_bo_update failed\n"); return ret; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 88681abc5f8c..12bad207bb0f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -806,7 +806,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (r) return r; - r = amdgpu_vm_bo_update(adev, fpriv->prt_va, false, NULL); + r = amdgpu_vm_bo_update(adev, fpriv->prt_va, false); if (r) return r; @@ -817,7 +817,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (amdgpu_mcbp || amdgpu_sriov_vf(adev)) { bo_va = fpriv->csa_va; BUG_ON(!bo_va); - r = amdgpu_vm_bo_update(adev, bo_va, false, NULL); + r = amdgpu_vm_bo_update(adev, bo_va, false); if (r) return r; @@ -836,7 +836,7 @@ static int amdgpu_cs_vm_handling(struct amdgpu_cs_parser *p) if (bo_va == NULL) continue; - r = amdgpu_vm_bo_update(adev, bo_va, false, NULL); + r = amdgpu_vm_bo_update(adev, bo_va, false); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index 57b74d35052f..89c6d6f1d4fa 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -612,7 +612,7 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev, if (operation == AMDGPU_VA_OP_MAP || operation == AMDGPU_VA_OP_REPLACE) { - r = amdgpu_vm_bo_update(adev, bo_va, false, NULL); + r = amdgpu_vm_bo_update(adev, bo_va, false); if (r) goto error; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index ecdcf652ba41..48f326609976 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -808,7 +808,6 @@ static void amdgpu_vm_tlb_seq_cb(struct dma_fence *fence, * @res: ttm_resource to map * @pages_addr: DMA addresses to use for mapping * @fence: optional resulting fence - * @table_freed: return true if page table is freed * * Fill in the page table entries between @start and @last. * @@ -823,8 +822,7 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, uint64_t flags, uint64_t offset, struct ttm_resource *res, dma_addr_t *pages_addr, - struct dma_fence **fence, - bool *table_freed) + struct dma_fence **fence) { struct amdgpu_vm_update_params params; struct amdgpu_vm_tlb_seq_cb *tlb_cb; @@ -938,9 +936,6 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, tlb_cb = NULL; } - if (table_freed) - *table_freed = *table_freed || params.table_freed; - error_free: kfree(tlb_cb); @@ -1000,7 +995,6 @@ void amdgpu_vm_get_memory(struct amdgpu_vm *vm, uint64_t *vram_mem, * @adev: amdgpu_device pointer * @bo_va: requested BO and VM object * @clear: if true clear the entries - * @table_freed: return true if page table is freed * * Fill in the page table entries for @bo_va. * @@ -1008,7 +1002,7 @@ void amdgpu_vm_get_memory(struct amdgpu_vm *vm, uint64_t *vram_mem, * 0 for success, -EINVAL for failure. */ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, - bool clear, bool *table_freed) + bool clear) { struct amdgpu_bo *bo = bo_va->base.bo; struct amdgpu_vm *vm = bo_va->base.vm; @@ -1087,7 +1081,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, resv, mapping->start, mapping->last, update_flags, mapping->offset, mem, - pages_addr, last_update, table_freed); + pages_addr, last_update); if (r) return r; } @@ -1281,7 +1275,7 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev, r = amdgpu_vm_bo_update_mapping(adev, adev, vm, false, false, resv, mapping->start, mapping->last, init_pte_value, - 0, NULL, NULL, &f, NULL); + 0, NULL, NULL, &f); amdgpu_vm_free_mapping(adev, vm, mapping, f); if (r) { dma_fence_put(f); @@ -1323,7 +1317,7 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev, list_for_each_entry_safe(bo_va, tmp, &vm->moved, base.vm_status) { /* Per VM BOs never need to bo cleared in the page tables */ - r = amdgpu_vm_bo_update(adev, bo_va, false, NULL); + r = amdgpu_vm_bo_update(adev, bo_va, false); if (r) return r; } @@ -1342,7 +1336,7 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev, else clear = true; - r = amdgpu_vm_bo_update(adev, bo_va, clear, NULL); + r = amdgpu_vm_bo_update(adev, bo_va, clear); if (r) return r; @@ -2526,8 +2520,7 @@ bool amdgpu_vm_handle_fault(struct amdgpu_device *adev, u32 pasid, } r = amdgpu_vm_bo_update_mapping(adev, adev, vm, true, false, NULL, addr, - addr, flags, value, NULL, NULL, NULL, - NULL); + addr, flags, value, NULL, NULL, NULL); if (r) goto error_unlock; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index 38a1eab1ff74..6b7682fe84f8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -410,10 +410,10 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, uint64_t flags, uint64_t offset, struct ttm_resource *res, dma_addr_t *pages_addr, - struct dma_fence **fence, bool *free_table); + struct dma_fence **fence); int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, - bool clear, bool *table_freed); + bool clear); bool amdgpu_vm_evictable(struct amdgpu_bo *bo); void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev, struct amdgpu_bo *bo, bool evicted); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index bf6354f539d8..27533f6057e0 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -1190,7 +1190,7 @@ svm_range_unmap_from_gpu(struct amdgpu_device *adev, struct amdgpu_vm *vm, return amdgpu_vm_bo_update_mapping(adev, adev, vm, false, true, NULL, start, last, init_pte_value, 0, - NULL, NULL, fence, NULL); + NULL, NULL, fence); } static int @@ -1282,8 +1282,7 @@ svm_range_map_to_gpu(struct kfd_process_device *pdd, struct svm_range *prange, prange->start + i, pte_flags, last_start - prange->start, NULL, dma_addr, - &vm->last_update, - NULL); + &vm->last_update); for (j = last_start - prange->start; j <= i; j++) dma_addr[j] |= last_domain; -- cgit v1.2.3 From 5d8539d2fbe39ce24e76dfe72b7de7093d3fac6d Mon Sep 17 00:00:00 2001 From: Darren Powell Date: Fri, 4 Mar 2022 23:45:25 -0500 Subject: amdgpu/pm: Add new hwmgr API function "emit_clock_levels" Extend commit 7f36948c92b2 ("amdgpu/pm: Implement new API function "emit" that accepts buffer base and write offset") Add new hwmgr API function "emit_clock_levels" - add member emit_clock_levels to pp_hwmgr_func - Implemented pp_dpm_emit_clock_levels - add pp_dpm_emit_clock_levels to pp_dpm_funcs Signed-off-by: Darren Powell Reviewed-by: Evan Quan Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c | 17 +++++++++++++++++ drivers/gpu/drm/amd/pm/powerplay/inc/hwmgr.h | 2 ++ 2 files changed, 19 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c index a2da46bf3985..dbed72c1e0c6 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c +++ b/drivers/gpu/drm/amd/pm/powerplay/amd_powerplay.c @@ -671,6 +671,22 @@ static int pp_dpm_force_clock_level(void *handle, return hwmgr->hwmgr_func->force_clock_level(hwmgr, type, mask); } +static int pp_dpm_emit_clock_levels(void *handle, + enum pp_clock_type type, + char *buf, + int *offset) +{ + struct pp_hwmgr *hwmgr = handle; + + if (!hwmgr || !hwmgr->pm_en) + return -EOPNOTSUPP; + + if (!hwmgr->hwmgr_func->emit_clock_levels) + return -ENOENT; + + return hwmgr->hwmgr_func->emit_clock_levels(hwmgr, type, buf, offset); +} + static int pp_dpm_print_clock_levels(void *handle, enum pp_clock_type type, char *buf) { @@ -1535,6 +1551,7 @@ static const struct amd_pm_funcs pp_dpm_funcs = { .get_pp_table = pp_dpm_get_pp_table, .set_pp_table = pp_dpm_set_pp_table, .force_clock_level = pp_dpm_force_clock_level, + .emit_clock_levels = pp_dpm_emit_clock_levels, .print_clock_levels = pp_dpm_print_clock_levels, .get_sclk_od = pp_dpm_get_sclk_od, .set_sclk_od = pp_dpm_set_sclk_od, diff --git a/drivers/gpu/drm/amd/pm/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/pm/powerplay/inc/hwmgr.h index 4f7f2f455301..27f8d0e0e6a8 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/inc/hwmgr.h +++ b/drivers/gpu/drm/amd/pm/powerplay/inc/hwmgr.h @@ -313,6 +313,8 @@ struct pp_hwmgr_func { int (*get_max_high_clocks)(struct pp_hwmgr *hwmgr, struct amd_pp_simple_clock_info *clocks); int (*power_off_asic)(struct pp_hwmgr *hwmgr); int (*force_clock_level)(struct pp_hwmgr *hwmgr, enum pp_clock_type type, uint32_t mask); + int (*emit_clock_levels)(struct pp_hwmgr *hwmgr, + enum pp_clock_type type, char *buf, int *offset); int (*print_clock_levels)(struct pp_hwmgr *hwmgr, enum pp_clock_type type, char *buf); int (*powergate_gfx)(struct pp_hwmgr *hwmgr, bool enable); int (*get_sclk_od)(struct pp_hwmgr *hwmgr); -- cgit v1.2.3 From a63e6b83c21424b6ecfca7ecbf605f9af75ae296 Mon Sep 17 00:00:00 2001 From: Darren Powell Date: Mon, 7 Mar 2022 21:40:35 -0500 Subject: amdgpu/pm: Implement emit_clk_levels for vega10 (v1) - implement emit_clk_levels for vega10, based on print_clk_levels, but using sysfs_emit rather than sprintf - modify local int vars to use uint32_t to match arg type of called functions - add return of error codes - refactor OD_XXX cases to return early with -EOPNOTSUPP if !(hwmgr->od_enabled) Signed-off-by: Darren Powell Reviewed-by: Evan Quan Signed-off-by: Alex Deucher --- .../gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c | 147 +++++++++++++++++++++ 1 file changed, 147 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c index 37324f2009ca..99bfe5efe171 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_hwmgr.c @@ -4625,6 +4625,152 @@ static int vega10_get_current_pcie_link_speed_level(struct pp_hwmgr *hwmgr) >> PSWUSP0_PCIE_LC_SPEED_CNTL__LC_CURRENT_DATA_RATE__SHIFT; } +static int vega10_emit_clock_levels(struct pp_hwmgr *hwmgr, + enum pp_clock_type type, char *buf, int *offset) +{ + struct vega10_hwmgr *data = hwmgr->backend; + struct vega10_single_dpm_table *sclk_table = &(data->dpm_table.gfx_table); + struct vega10_single_dpm_table *mclk_table = &(data->dpm_table.mem_table); + struct vega10_single_dpm_table *soc_table = &(data->dpm_table.soc_table); + struct vega10_single_dpm_table *dcef_table = &(data->dpm_table.dcef_table); + struct vega10_odn_clock_voltage_dependency_table *podn_vdd_dep = NULL; + uint32_t gen_speed, lane_width, current_gen_speed, current_lane_width; + PPTable_t *pptable = &(data->smc_state_table.pp_table); + + uint32_t i, now, count = 0; + int ret = 0; + + switch (type) { + case PP_SCLK: + if (data->registry_data.sclk_dpm_key_disabled) + return -EOPNOTSUPP; + + ret = smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetCurrentGfxclkIndex, &now); + if (unlikely(ret != 0)) + return ret; + + if (hwmgr->pp_one_vf && + (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK)) + count = 5; + else + count = sclk_table->count; + for (i = 0; i < count; i++) + *offset += sysfs_emit_at(buf, *offset, "%d: %uMhz %s\n", + i, sclk_table->dpm_levels[i].value / 100, + (i == now) ? "*" : ""); + break; + case PP_MCLK: + if (data->registry_data.mclk_dpm_key_disabled) + return -EOPNOTSUPP; + + ret = smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetCurrentUclkIndex, &now); + if (unlikely(ret != 0)) + return ret; + + for (i = 0; i < mclk_table->count; i++) + *offset += sysfs_emit_at(buf, *offset, "%d: %uMhz %s\n", + i, mclk_table->dpm_levels[i].value / 100, + (i == now) ? "*" : ""); + break; + case PP_SOCCLK: + if (data->registry_data.socclk_dpm_key_disabled) + return -EOPNOTSUPP; + + ret = smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetCurrentSocclkIndex, &now); + if (unlikely(ret != 0)) + return ret; + + for (i = 0; i < soc_table->count; i++) + *offset += sysfs_emit_at(buf, *offset, "%d: %uMhz %s\n", + i, soc_table->dpm_levels[i].value / 100, + (i == now) ? "*" : ""); + break; + case PP_DCEFCLK: + if (data->registry_data.dcefclk_dpm_key_disabled) + return -EOPNOTSUPP; + + ret = smum_send_msg_to_smc_with_parameter(hwmgr, + PPSMC_MSG_GetClockFreqMHz, + CLK_DCEFCLK, &now); + if (unlikely(ret != 0)) + return ret; + + for (i = 0; i < dcef_table->count; i++) + *offset += sysfs_emit_at(buf, *offset, "%d: %uMhz %s\n", + i, dcef_table->dpm_levels[i].value / 100, + (dcef_table->dpm_levels[i].value / 100 == now) ? + "*" : ""); + break; + case PP_PCIE: + current_gen_speed = + vega10_get_current_pcie_link_speed_level(hwmgr); + current_lane_width = + vega10_get_current_pcie_link_width_level(hwmgr); + for (i = 0; i < NUM_LINK_LEVELS; i++) { + gen_speed = pptable->PcieGenSpeed[i]; + lane_width = pptable->PcieLaneCount[i]; + + *offset += sysfs_emit_at(buf, *offset, "%d: %s %s %s\n", i, + (gen_speed == 0) ? "2.5GT/s," : + (gen_speed == 1) ? "5.0GT/s," : + (gen_speed == 2) ? "8.0GT/s," : + (gen_speed == 3) ? "16.0GT/s," : "", + (lane_width == 1) ? "x1" : + (lane_width == 2) ? "x2" : + (lane_width == 3) ? "x4" : + (lane_width == 4) ? "x8" : + (lane_width == 5) ? "x12" : + (lane_width == 6) ? "x16" : "", + (current_gen_speed == gen_speed) && + (current_lane_width == lane_width) ? + "*" : ""); + } + break; + + case OD_SCLK: + if (!hwmgr->od_enabled) + return -EOPNOTSUPP; + + *offset += sysfs_emit_at(buf, *offset, "%s:\n", "OD_SCLK"); + podn_vdd_dep = &data->odn_dpm_table.vdd_dep_on_sclk; + for (i = 0; i < podn_vdd_dep->count; i++) + *offset += sysfs_emit_at(buf, *offset, "%d: %10uMhz %10umV\n", + i, podn_vdd_dep->entries[i].clk / 100, + podn_vdd_dep->entries[i].vddc); + break; + case OD_MCLK: + if (!hwmgr->od_enabled) + return -EOPNOTSUPP; + + *offset += sysfs_emit_at(buf, *offset, "%s:\n", "OD_MCLK"); + podn_vdd_dep = &data->odn_dpm_table.vdd_dep_on_mclk; + for (i = 0; i < podn_vdd_dep->count; i++) + *offset += sysfs_emit_at(buf, *offset, "%d: %10uMhz %10umV\n", + i, podn_vdd_dep->entries[i].clk/100, + podn_vdd_dep->entries[i].vddc); + break; + case OD_RANGE: + if (!hwmgr->od_enabled) + return -EOPNOTSUPP; + + *offset += sysfs_emit_at(buf, *offset, "%s:\n", "OD_RANGE"); + *offset += sysfs_emit_at(buf, *offset, "SCLK: %7uMHz %10uMHz\n", + data->golden_dpm_table.gfx_table.dpm_levels[0].value/100, + hwmgr->platform_descriptor.overdriveLimit.engineClock/100); + *offset += sysfs_emit_at(buf, *offset, "MCLK: %7uMHz %10uMHz\n", + data->golden_dpm_table.mem_table.dpm_levels[0].value/100, + hwmgr->platform_descriptor.overdriveLimit.memoryClock/100); + *offset += sysfs_emit_at(buf, *offset, "VDDC: %7umV %11umV\n", + data->odn_dpm_table.min_vddc, + data->odn_dpm_table.max_vddc); + break; + default: + ret = -ENOENT; + break; + } + return ret; +} + static int vega10_print_clock_levels(struct pp_hwmgr *hwmgr, enum pp_clock_type type, char *buf) { @@ -5559,6 +5705,7 @@ static const struct pp_hwmgr_func vega10_hwmgr_funcs = { .set_watermarks_for_clocks_ranges = vega10_set_watermarks_for_clocks_ranges, .display_clock_voltage_request = vega10_display_clock_voltage_request, .force_clock_level = vega10_force_clock_level, + .emit_clock_levels = vega10_emit_clock_levels, .print_clock_levels = vega10_print_clock_levels, .display_config_changed = vega10_display_configuration_changed_task, .powergate_uvd = vega10_power_gate_uvd, -- cgit v1.2.3 From e3026a057f9675884840979f17cd230e52685792 Mon Sep 17 00:00:00 2001 From: Boyuan Zhang Date: Thu, 24 Mar 2022 12:19:06 -0400 Subject: drm/amdgpu/vcn3: send smu interface type For VCN FW to detect ASIC type, in order to use different mailbox registers. V2: simplify codes and fix format issue. Signed-off-by: Boyuan Zhang Acked-by Huang Rui Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h | 7 +++++++ drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c | 5 +++++ 2 files changed, 12 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h index e2fde88aaf5e..f06fb7f882e2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h @@ -159,6 +159,7 @@ #define AMDGPU_VCN_MULTI_QUEUE_FLAG (1 << 8) #define AMDGPU_VCN_SW_RING_FLAG (1 << 9) #define AMDGPU_VCN_FW_LOGGING_FLAG (1 << 10) +#define AMDGPU_VCN_SMU_VERSION_INFO_FLAG (1 << 11) #define AMDGPU_VCN_IB_FLAG_DECODE_BUFFER 0x00000001 #define AMDGPU_VCN_CMD_FLAG_MSG_BUFFER 0x00000001 @@ -279,6 +280,11 @@ struct amdgpu_fw_shared_fw_logging { uint32_t size; }; +struct amdgpu_fw_shared_smu_interface_info { + uint8_t smu_interface_type; + uint8_t padding[3]; +}; + struct amdgpu_fw_shared { uint32_t present_flag_0; uint8_t pad[44]; @@ -287,6 +293,7 @@ struct amdgpu_fw_shared { struct amdgpu_fw_shared_multi_queue multi_queue; struct amdgpu_fw_shared_sw_ring sw_ring; struct amdgpu_fw_shared_fw_logging fw_log; + struct amdgpu_fw_shared_smu_interface_info smu_interface_info; }; struct amdgpu_vcn_fwlog { diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c index 0d590183328f..e1cca0a10653 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c @@ -219,6 +219,11 @@ static int vcn_v3_0_sw_init(void *handle) cpu_to_le32(AMDGPU_VCN_MULTI_QUEUE_FLAG) | cpu_to_le32(AMDGPU_VCN_FW_SHARED_FLAG_0_RB); fw_shared->sw_ring.is_enabled = cpu_to_le32(DEC_SW_RING_ENABLED); + fw_shared->present_flag_0 |= AMDGPU_VCN_SMU_VERSION_INFO_FLAG; + if (adev->ip_versions[UVD_HWIP][0] == IP_VERSION(3, 1, 2)) + fw_shared->smu_interface_info.smu_interface_type = 2; + else if (adev->ip_versions[UVD_HWIP][0] == IP_VERSION(3, 1, 1)) + fw_shared->smu_interface_info.smu_interface_type = 1; if (amdgpu_vcnfw_log) amdgpu_vcn_fwlog_init(&adev->vcn.inst[i]); -- cgit v1.2.3 From 8ecad8d66d7f03ccc8fa7b4921d45041e9b77680 Mon Sep 17 00:00:00 2001 From: Lijo Lazar Date: Fri, 25 Mar 2022 13:19:36 +0530 Subject: drm/amd/pm: Check feature support using IP version Instead of ASIC type, use GC and MP1 IP versions for feature support checks. Signed-off-by: Lijo Lazar Reviewed-by: Kevin Wang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/amdgpu_pm.c | 73 +++++++++++++++++++++----------------- 1 file changed, 40 insertions(+), 33 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c index 5cd67ddf8495..9ce597ded31d 100644 --- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c @@ -1954,8 +1954,9 @@ static int default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_ uint32_t mask, enum amdgpu_device_attr_states *states) { struct device_attribute *dev_attr = &attr->dev_attr; + uint32_t mp1_ver = adev->ip_versions[MP1_HWIP][0]; + uint32_t gc_ver = adev->ip_versions[GC_HWIP][0]; const char *attr_name = dev_attr->attr.name; - enum amd_asic_type asic_type = adev->asic_type; if (!(attr->flags & mask)) { *states = ATTR_STATE_UNSUPPORTED; @@ -1965,53 +1966,55 @@ static int default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_ #define DEVICE_ATTR_IS(_name) (!strcmp(attr_name, #_name)) if (DEVICE_ATTR_IS(pp_dpm_socclk)) { - if (asic_type < CHIP_VEGA10) + if (gc_ver < IP_VERSION(9, 0, 0)) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pp_dpm_dcefclk)) { - if (asic_type < CHIP_VEGA10 || - asic_type == CHIP_ARCTURUS || - asic_type == CHIP_ALDEBARAN) + if (gc_ver < IP_VERSION(9, 0, 0) || + gc_ver == IP_VERSION(9, 4, 1) || + gc_ver == IP_VERSION(9, 4, 2)) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pp_dpm_fclk)) { - if (asic_type < CHIP_VEGA20) + if (mp1_ver < IP_VERSION(10, 0, 0)) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pp_od_clk_voltage)) { *states = ATTR_STATE_UNSUPPORTED; if (amdgpu_dpm_is_overdrive_supported(adev)) *states = ATTR_STATE_SUPPORTED; } else if (DEVICE_ATTR_IS(mem_busy_percent)) { - if (adev->flags & AMD_IS_APU || asic_type == CHIP_VEGA10) + if (adev->flags & AMD_IS_APU || gc_ver == IP_VERSION(9, 0, 1)) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pcie_bw)) { /* PCIe Perf counters won't work on APU nodes */ if (adev->flags & AMD_IS_APU) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(unique_id)) { - if (asic_type != CHIP_VEGA10 && - asic_type != CHIP_VEGA20 && - asic_type != CHIP_ARCTURUS && - asic_type != CHIP_ALDEBARAN) + if (gc_ver != IP_VERSION(9, 0, 1) && + gc_ver != IP_VERSION(9, 4, 0) && + gc_ver != IP_VERSION(9, 4, 1) && + gc_ver != IP_VERSION(9, 4, 2)) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pp_features)) { - if (adev->flags & AMD_IS_APU || asic_type < CHIP_VEGA10) + if (adev->flags & AMD_IS_APU || gc_ver < IP_VERSION(9, 0, 0)) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(gpu_metrics)) { - if (asic_type < CHIP_VEGA12) + if (gc_ver < IP_VERSION(9, 1, 0)) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pp_dpm_vclk)) { - if (!(asic_type == CHIP_VANGOGH || asic_type == CHIP_SIENNA_CICHLID)) + if (!(gc_ver == IP_VERSION(10, 3, 1) || + gc_ver == IP_VERSION(10, 3, 0))) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pp_dpm_dclk)) { - if (!(asic_type == CHIP_VANGOGH || asic_type == CHIP_SIENNA_CICHLID)) + if (!(gc_ver == IP_VERSION(10, 3, 1) || + gc_ver == IP_VERSION(10, 3, 0))) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pp_power_profile_mode)) { if (amdgpu_dpm_get_power_profile_mode(adev, NULL) == -EOPNOTSUPP) *states = ATTR_STATE_UNSUPPORTED; } - switch (asic_type) { - case CHIP_ARCTURUS: - case CHIP_ALDEBARAN: + switch (gc_ver) { + case IP_VERSION(9, 4, 1): + case IP_VERSION(9, 4, 2): /* the Mi series card does not support standalone mclk/socclk/fclk level setting */ if (DEVICE_ATTR_IS(pp_dpm_mclk) || DEVICE_ATTR_IS(pp_dpm_socclk) || @@ -2026,7 +2029,7 @@ static int default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_ if (DEVICE_ATTR_IS(pp_dpm_dcefclk)) { /* SMU MP1 does not support dcefclk level setting */ - if (asic_type >= CHIP_NAVI10) { + if (gc_ver >= IP_VERSION(10, 0, 0)) { dev_attr->attr.mode &= ~S_IWUGO; dev_attr->store = NULL; } @@ -2864,8 +2867,9 @@ static ssize_t amdgpu_hwmon_show_power_label(struct device *dev, char *buf) { struct amdgpu_device *adev = dev_get_drvdata(dev); + uint32_t gc_ver = adev->ip_versions[GC_HWIP][0]; - if (adev->asic_type == CHIP_VANGOGH) + if (gc_ver == IP_VERSION(10, 3, 1)) return sysfs_emit(buf, "%s\n", to_sensor_dev_attr(attr)->index == PP_PWR_TYPE_FAST ? "fastPPT" : "slowPPT"); @@ -3177,6 +3181,7 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj, struct device *dev = kobj_to_dev(kobj); struct amdgpu_device *adev = dev_get_drvdata(dev); umode_t effective_mode = attr->mode; + uint32_t gc_ver = adev->ip_versions[GC_HWIP][0]; /* under multi-vf mode, the hwmon attributes are all not supported */ if (amdgpu_sriov_vf(adev) && !amdgpu_sriov_is_pp_one_vf(adev)) @@ -3245,18 +3250,18 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj, attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr)) /* can't manage state */ effective_mode &= ~S_IWUSR; + /* not implemented yet for GC 10.3.1 APUs */ if (((adev->family == AMDGPU_FAMILY_SI) || - ((adev->flags & AMD_IS_APU) && - (adev->asic_type != CHIP_VANGOGH))) && /* not implemented yet */ + ((adev->flags & AMD_IS_APU) && (gc_ver != IP_VERSION(10, 3, 1)))) && (attr == &sensor_dev_attr_power1_cap_max.dev_attr.attr || - attr == &sensor_dev_attr_power1_cap_min.dev_attr.attr|| + attr == &sensor_dev_attr_power1_cap_min.dev_attr.attr || attr == &sensor_dev_attr_power1_cap.dev_attr.attr || attr == &sensor_dev_attr_power1_cap_default.dev_attr.attr)) return 0; + /* not implemented yet for APUs having <= GC 9.3.0 */ if (((adev->family == AMDGPU_FAMILY_SI) || - ((adev->flags & AMD_IS_APU) && - (adev->asic_type < CHIP_RENOIR))) && /* not implemented yet */ + ((adev->flags & AMD_IS_APU) && (gc_ver < IP_VERSION(9, 3, 0)))) && (attr == &sensor_dev_attr_power1_average.dev_attr.attr)) return 0; @@ -3294,8 +3299,7 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj, return 0; /* only SOC15 dGPUs support hotspot and mem temperatures */ - if (((adev->flags & AMD_IS_APU) || - adev->asic_type < CHIP_VEGA10) && + if (((adev->flags & AMD_IS_APU) || gc_ver < IP_VERSION(9, 0, 0)) && (attr == &sensor_dev_attr_temp2_crit.dev_attr.attr || attr == &sensor_dev_attr_temp2_crit_hyst.dev_attr.attr || attr == &sensor_dev_attr_temp3_crit.dev_attr.attr || @@ -3310,13 +3314,13 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj, return 0; /* only Vangogh has fast PPT limit and power labels */ - if (!(adev->asic_type == CHIP_VANGOGH) && + if (!(gc_ver == IP_VERSION(10, 3, 1)) && (attr == &sensor_dev_attr_power2_average.dev_attr.attr || - attr == &sensor_dev_attr_power2_cap_max.dev_attr.attr || + attr == &sensor_dev_attr_power2_cap_max.dev_attr.attr || attr == &sensor_dev_attr_power2_cap_min.dev_attr.attr || - attr == &sensor_dev_attr_power2_cap.dev_attr.attr || - attr == &sensor_dev_attr_power2_cap_default.dev_attr.attr || - attr == &sensor_dev_attr_power2_label.dev_attr.attr)) + attr == &sensor_dev_attr_power2_cap.dev_attr.attr || + attr == &sensor_dev_attr_power2_cap_default.dev_attr.attr || + attr == &sensor_dev_attr_power2_label.dev_attr.attr)) return 0; return effective_mode; @@ -3421,6 +3425,8 @@ static void amdgpu_debugfs_prints_cpu_info(struct seq_file *m, static int amdgpu_debugfs_pm_info_pp(struct seq_file *m, struct amdgpu_device *adev) { + uint32_t mp1_ver = adev->ip_versions[MP1_HWIP][0]; + uint32_t gc_ver = adev->ip_versions[GC_HWIP][0]; uint32_t value; uint64_t value64 = 0; uint32_t query = 0; @@ -3467,7 +3473,8 @@ static int amdgpu_debugfs_pm_info_pp(struct seq_file *m, struct amdgpu_device *a if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_ENABLED_SMC_FEATURES_MASK, (void *)&value64, &size)) seq_printf(m, "SMC Feature Mask: 0x%016llx\n", value64); - if (adev->asic_type > CHIP_VEGA20) { + /* ASICs greater than CHIP_VEGA20 supports these sensors */ + if (gc_ver != IP_VERSION(9, 4, 0) && mp1_ver > IP_VERSION(9, 0, 0)) { /* VCN clocks */ if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VCN_POWER_STATE, (void *)&value, &size)) { if (!value) { -- cgit v1.2.3 From 9f07550b3c8a79cbb996a368f699c9eb619d8322 Mon Sep 17 00:00:00 2001 From: Sean Paul Date: Fri, 25 Mar 2022 00:06:19 +0000 Subject: drm/amdgpu: Re-classify some log messages in commit path MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit ATOMIC and DRIVER log categories do not typically contain per-frame log messages. This patch re-classifies some messages in amd to chattier categories to keep ATOMIC/DRIVER quiet. Acked-by: Christian König Reviewed-by: Harry Wentland Signed-off-by: Sean Paul Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_display.c | 5 +++-- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 8 ++++---- 2 files changed, 7 insertions(+), 6 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c index ee21b242ccf6..787724166952 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c @@ -118,8 +118,9 @@ static void amdgpu_display_flip_work_func(struct work_struct *__work) spin_unlock_irqrestore(&crtc->dev->event_lock, flags); - DRM_DEBUG_DRIVER("crtc:%d[%p], pflip_stat:AMDGPU_FLIP_SUBMITTED, work: %p,\n", - amdgpu_crtc->crtc_id, amdgpu_crtc, work); + drm_dbg_vbl(adev_to_drm(adev), + "crtc:%d[%p], pflip_stat:AMDGPU_FLIP_SUBMITTED, work: %p,\n", + amdgpu_crtc->crtc_id, amdgpu_crtc, work); } 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 b30656959fd8..45d130f86114 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -9248,7 +9248,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, &bundle->flip_addrs[planes_count].address, afb->tmz_surface, false); - DRM_DEBUG_ATOMIC("plane: id=%d dcc_en=%d\n", + drm_dbg_state(state->dev, "plane: id=%d dcc_en=%d\n", new_plane_state->plane->index, bundle->plane_infos[planes_count].dcc.enable); @@ -9282,7 +9282,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state, dc_plane, bundle->flip_addrs[planes_count].flip_timestamp_in_us); - DRM_DEBUG_ATOMIC("%s Flipping to hi: 0x%x, low: 0x%x\n", + drm_dbg_state(state->dev, "%s Flipping to hi: 0x%x, low: 0x%x\n", __func__, bundle->flip_addrs[planes_count].address.grph.addr.high_part, bundle->flip_addrs[planes_count].address.grph.addr.low_part); @@ -9624,7 +9624,7 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) dm_new_crtc_state = to_dm_crtc_state(new_crtc_state); dm_old_crtc_state = to_dm_crtc_state(old_crtc_state); - DRM_DEBUG_ATOMIC( + drm_dbg_state(state->dev, "amdgpu_crtc id:%d crtc_state_flags: enable:%d, active:%d, " "planes_changed:%d, mode_changed:%d,active_changed:%d," "connectors_changed:%d\n", @@ -10328,7 +10328,7 @@ static int dm_update_crtc_state(struct amdgpu_display_manager *dm, if (!drm_atomic_crtc_needs_modeset(new_crtc_state)) goto skip_modeset; - DRM_DEBUG_ATOMIC( + drm_dbg_state(state->dev, "amdgpu_crtc id:%d crtc_state_flags: enable:%d, active:%d, " "planes_changed:%d, mode_changed:%d,active_changed:%d," "connectors_changed:%d\n", -- cgit v1.2.3 From 425d7a87e54ee358f580eaf10cf28dc95f7121c1 Mon Sep 17 00:00:00 2001 From: tiancyin Date: Sun, 27 Mar 2022 19:07:13 +0800 Subject: drm/amd/vcn: fix an error msg on vcn 3.0 Some video card has more than one vcn instance, passing 0 to vcn_v3_0_pause_dpg_mode is incorrect. Error msg: Register(1) [mmUVD_POWER_STATUS] failed to reach value 0x00000001 != 0x00000002 Reviewed-by: James Zhu Signed-off-by: tiancyin Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c index e1cca0a10653..cb5f0a12333f 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c @@ -1488,7 +1488,7 @@ static int vcn_v3_0_stop_dpg_mode(struct amdgpu_device *adev, int inst_idx) struct dpg_pause_state state = {.fw_based = VCN_DPG_STATE__UNPAUSE}; uint32_t tmp; - vcn_v3_0_pause_dpg_mode(adev, 0, &state); + vcn_v3_0_pause_dpg_mode(adev, inst_idx, &state); /* Wait for power status to be 1 */ SOC15_WAIT_ON_RREG(VCN, inst_idx, mmUVD_POWER_STATUS, 1, -- cgit v1.2.3 From a3d63c62bdf06c5a3f8a71c207f13b26fc6030f5 Mon Sep 17 00:00:00 2001 From: Mohammad Zafar Ziya Date: Thu, 17 Mar 2022 09:45:42 +0800 Subject: drm/amdgpu: Add vcn and jpeg ras support flag Add vcn and jpeg ras support options V2: vcn and jpeg ras flag enabled for aldebaran asic only V3: vcn and jpeg ras flag disabled for error counter query Generic poison query interface added VCN and JPEG ras enabled based on IP version check V4: vcn and jpeg ras flag moved under ecc flag for dGPU Signed-off-by: Mohammad Zafar Ziya Reviewed-by: Hawking Zhang Reviewed-by: Tao Zhou Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 9 +++++++++ drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h | 2 ++ 2 files changed, 11 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index 424c22a841f4..ec709997c9c7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -66,6 +66,8 @@ const char *ras_block_string[] = { "mp1", "fuse", "mca", + "vcn", + "jpeg", }; const char *ras_mca_block_string[] = { @@ -2205,6 +2207,13 @@ static void amdgpu_ras_check_supported(struct amdgpu_device *adev) dev_info(adev->dev, "SRAM ECC is active.\n"); adev->ras_hw_enabled |= ~(1 << AMDGPU_RAS_BLOCK__UMC | 1 << AMDGPU_RAS_BLOCK__DF); + + if (adev->ip_versions[VCN_HWIP][0] == IP_VERSION(2, 6, 0)) + adev->ras_hw_enabled |= (1 << AMDGPU_RAS_BLOCK__VCN | + 1 << AMDGPU_RAS_BLOCK__JPEG); + else + adev->ras_hw_enabled &= ~(1 << AMDGPU_RAS_BLOCK__VCN | + 1 << AMDGPU_RAS_BLOCK__JPEG); } else { dev_info(adev->dev, "SRAM ECC is not presented.\n"); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h index 9314fde81e68..1e1a3b736859 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h @@ -49,6 +49,8 @@ enum amdgpu_ras_block { AMDGPU_RAS_BLOCK__MP1, AMDGPU_RAS_BLOCK__FUSE, AMDGPU_RAS_BLOCK__MCA, + AMDGPU_RAS_BLOCK__VCN, + AMDGPU_RAS_BLOCK__JPEG, AMDGPU_RAS_BLOCK__LAST }; -- cgit v1.2.3 From 60fce7417fd6c0fef34ca1661bb17d63cf6db30b Mon Sep 17 00:00:00 2001 From: Mohammad Zafar Ziya Date: Thu, 17 Mar 2022 09:49:19 +0800 Subject: drm/amdgpu/vcn: Add vcn ras support VCN block ras feature support addition V2: default ras callback removed Signed-off-by: Mohammad Zafar Ziya Reviewed-by: Hawking Zhang Reviewed-by: Tao Zhou Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h index f06fb7f882e2..fb39065a96bd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h @@ -24,6 +24,8 @@ #ifndef __AMDGPU_VCN_H__ #define __AMDGPU_VCN_H__ +#include "amdgpu_ras.h" + #define AMDGPU_VCN_STACK_SIZE (128*1024) #define AMDGPU_VCN_CONTEXT_SIZE (512*1024) @@ -233,6 +235,10 @@ struct amdgpu_vcn_inst { struct amdgpu_vcn_fw_shared fw_shared; }; +struct amdgpu_vcn_ras { + struct amdgpu_ras_block_object ras_block; +}; + struct amdgpu_vcn { unsigned fw_version; struct delayed_work idle_work; @@ -252,6 +258,9 @@ struct amdgpu_vcn { unsigned harvest_config; int (*pause_dpg_mode)(struct amdgpu_device *adev, int inst_idx, struct dpg_pause_state *new_state); + + struct ras_common_if *ras_if; + struct amdgpu_vcn_ras *ras; }; struct amdgpu_fw_shared_rb_ptrs_struct { -- cgit v1.2.3 From edd08fa137d39b1100296049c86c514682dbf378 Mon Sep 17 00:00:00 2001 From: Mohammad Zafar Ziya Date: Thu, 17 Mar 2022 09:53:03 +0800 Subject: drm/amdgpu/jpeg: Add jpeg block ras support Ras support addition for JPEG block V2: removed default callback Signed-off-by: Mohammad Zafar Ziya Reviewed-by: Hawking Zhang Reviewed-by: Tao Zhou Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.h | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.h index 55fbff2be761..b6c7fb00e05a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_jpeg.h @@ -24,6 +24,8 @@ #ifndef __AMDGPU_JPEG_H__ #define __AMDGPU_JPEG_H__ +#include "amdgpu_ras.h" + #define AMDGPU_MAX_JPEG_INSTANCES 2 #define AMDGPU_JPEG_HARVEST_JPEG0 (1 << 0) @@ -39,6 +41,10 @@ struct amdgpu_jpeg_inst { struct amdgpu_jpeg_reg external; }; +struct amdgpu_jpeg_ras { + struct amdgpu_ras_block_object ras_block; +}; + struct amdgpu_jpeg { uint8_t num_jpeg_inst; struct amdgpu_jpeg_inst inst[AMDGPU_MAX_JPEG_INSTANCES]; @@ -48,6 +54,8 @@ struct amdgpu_jpeg { enum amd_powergating_state cur_state; struct mutex jpeg_pg_lock; atomic_t total_submission_cnt; + struct ras_common_if *ras_if; + struct amdgpu_jpeg_ras *ras; }; int amdgpu_jpeg_sw_init(struct amdgpu_device *adev); -- cgit v1.2.3 From f0a339a8c0963abc80640975585df469e27ba3df Mon Sep 17 00:00:00 2001 From: Mohammad Zafar Ziya Date: Wed, 23 Mar 2022 10:54:15 +0800 Subject: drm/amdgpu/vcn: Add vcn and jpeg ver 2.6 ras register definition Adding vcn and jpeg ver 2.6 ras register definition Signed-off-by: Mohammad Zafar Ziya Reviewed-by: Hawking Zhang Reviewed-by: Tao Zhou Signed-off-by: Alex Deucher --- .../drm/amd/include/asic_reg/vcn/vcn_2_5_offset.h | 13 ++++++++++++ .../drm/amd/include/asic_reg/vcn/vcn_2_5_sh_mask.h | 24 ++++++++++++++++++++++ 2 files changed, 37 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_2_5_offset.h b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_2_5_offset.h index 90350f46a0c4..363d2139cea2 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_2_5_offset.h +++ b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_2_5_offset.h @@ -988,4 +988,17 @@ #define mmMDM_WIG_PIPE_BUSY_BASE_IDX 1 +/* VCN 2_6_0 regs */ +#define mmUVD_RAS_VCPU_VCODEC_STATUS 0x0057 +#define mmUVD_RAS_VCPU_VCODEC_STATUS_BASE_IDX 1 +#define mmUVD_RAS_MMSCH_FATAL_ERROR 0x0058 +#define mmUVD_RAS_MMSCH_FATAL_ERROR_BASE_IDX 1 + + +/* JPEG 2_6_0 regs */ +#define mmUVD_RAS_JPEG0_STATUS 0x0059 +#define mmUVD_RAS_JPEG0_STATUS_BASE_IDX 1 +#define mmUVD_RAS_JPEG1_STATUS 0x005a +#define mmUVD_RAS_JPEG1_STATUS_BASE_IDX 1 + #endif diff --git a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_2_5_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_2_5_sh_mask.h index c41c59c30006..8de883b76d90 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_2_5_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_2_5_sh_mask.h @@ -3606,4 +3606,28 @@ #define UVD_LMI_CRC3__CRC32_MASK 0xFFFFFFFFL +/* VCN 2_6_0 UVD_RAS_VCPU_VCODEC_STATUS */ +#define UVD_RAS_VCPU_VCODEC_STATUS__POISONED_VF__SHIFT 0x0 +#define UVD_RAS_VCPU_VCODEC_STATUS__POISONED_PF__SHIFT 0x1f +#define UVD_RAS_VCPU_VCODEC_STATUS__POISONED_VF_MASK 0x7FFFFFFFL +#define UVD_RAS_VCPU_VCODEC_STATUS__POISONED_PF_MASK 0x80000000L + +/* VCN 2_6_0 UVD_RAS_MMSCH_FATAL_ERROR */ +#define UVD_RAS_MMSCH_FATAL_ERROR__POISONED_VF__SHIFT 0x0 +#define UVD_RAS_MMSCH_FATAL_ERROR__POISONED_PF__SHIFT 0x1f +#define UVD_RAS_MMSCH_FATAL_ERROR__POISONED_VF_MASK 0x7FFFFFFFL +#define UVD_RAS_MMSCH_FATAL_ERROR__POISONED_PF_MASK 0x80000000L + +/* JPEG 2_6_0 UVD_RAS_JPEG0_STATUS */ +#define UVD_RAS_JPEG0_STATUS__POISONED_VF__SHIFT 0x0 +#define UVD_RAS_JPEG0_STATUS__POISONED_PF__SHIFT 0x1f +#define UVD_RAS_JPEG0_STATUS__POISONED_VF_MASK 0x7FFFFFFFL +#define UVD_RAS_JPEG0_STATUS__POISONED_PF_MASK 0x80000000L + +/* JPEG 2_6_0 UVD_RAS_JPEG1_STATUS */ +#define UVD_RAS_JPEG1_STATUS__POISONED_VF__SHIFT 0x0 +#define UVD_RAS_JPEG1_STATUS__POISONED_PF__SHIFT 0x1f +#define UVD_RAS_JPEG1_STATUS__POISONED_VF_MASK 0x7FFFFFFFL +#define UVD_RAS_JPEG1_STATUS__POISONED_PF_MASK 0x80000000L + #endif -- cgit v1.2.3 From c543dcbe4237e03b23fa40e0fba979cfd8514954 Mon Sep 17 00:00:00 2001 From: Mohammad Zafar Ziya Date: Wed, 23 Mar 2022 12:06:49 +0800 Subject: drm/amdgpu/vcn: Add VCN ras error query support RAS error query support addition for VCN 2.6 V2: removed unused option and corrected comment format Moved the register definition under header file V3: poison query status check added. Removed error query interface V4: MMSCH poison check option removed, return true/false refactored. Signed-off-by: Mohammad Zafar Ziya Reviewed-by: Hawking Zhang Reviewed-by: Tao Zhou Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h | 1 + drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c | 71 +++++++++++++++++++++++++++++++++ drivers/gpu/drm/amd/amdgpu/vcn_v2_5.h | 6 +++ 3 files changed, 78 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h index 1e1a3b736859..606df8869b89 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h @@ -508,6 +508,7 @@ struct amdgpu_ras_block_hw_ops { void (*query_ras_error_address)(struct amdgpu_device *adev, void *ras_error_status); void (*reset_ras_error_count)(struct amdgpu_device *adev); void (*reset_ras_error_status)(struct amdgpu_device *adev); + bool (*query_poison_status)(struct amdgpu_device *adev); }; /* work flow diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c index 1bf672966a62..3e1de8c8b372 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c @@ -31,6 +31,7 @@ #include "soc15d.h" #include "vcn_v2_0.h" #include "mmsch_v1_0.h" +#include "vcn_v2_5.h" #include "vcn/vcn_2_5_offset.h" #include "vcn/vcn_2_5_sh_mask.h" @@ -59,6 +60,7 @@ static int vcn_v2_5_set_powergating_state(void *handle, static int vcn_v2_5_pause_dpg_mode(struct amdgpu_device *adev, int inst_idx, struct dpg_pause_state *new_state); static int vcn_v2_5_sriov_start(struct amdgpu_device *adev); +static void vcn_v2_5_set_ras_funcs(struct amdgpu_device *adev); static int amdgpu_ih_clientid_vcns[] = { SOC15_IH_CLIENTID_VCN, @@ -100,6 +102,7 @@ static int vcn_v2_5_early_init(void *handle) vcn_v2_5_set_dec_ring_funcs(adev); vcn_v2_5_set_enc_ring_funcs(adev); vcn_v2_5_set_irq_funcs(adev); + vcn_v2_5_set_ras_funcs(adev); return 0; } @@ -1932,3 +1935,71 @@ const struct amdgpu_ip_block_version vcn_v2_6_ip_block = .rev = 0, .funcs = &vcn_v2_6_ip_funcs, }; + +static uint32_t vcn_v2_6_query_poison_by_instance(struct amdgpu_device *adev, + uint32_t instance, uint32_t sub_block) +{ + uint32_t poison_stat = 0, reg_value = 0; + + switch (sub_block) { + case AMDGPU_VCN_V2_6_VCPU_VCODEC: + reg_value = RREG32_SOC15(VCN, instance, mmUVD_RAS_VCPU_VCODEC_STATUS); + poison_stat = REG_GET_FIELD(reg_value, UVD_RAS_VCPU_VCODEC_STATUS, POISONED_PF); + break; + default: + break; + }; + + if (poison_stat) + dev_info(adev->dev, "Poison detected in VCN%d, sub_block%d\n", + instance, sub_block); + + return poison_stat; +} + +static bool vcn_v2_6_query_poison_status(struct amdgpu_device *adev) +{ + uint32_t inst, sub; + uint32_t poison_stat = 0; + + for (inst = 0; inst < adev->vcn.num_vcn_inst; inst++) + for (sub = 0; sub < AMDGPU_VCN_V2_6_MAX_SUB_BLOCK; sub++) + poison_stat += + vcn_v2_6_query_poison_by_instance(adev, inst, sub); + + return !!poison_stat; +} + +const struct amdgpu_ras_block_hw_ops vcn_v2_6_ras_hw_ops = { + .query_poison_status = vcn_v2_6_query_poison_status, +}; + +static struct amdgpu_vcn_ras vcn_v2_6_ras = { + .ras_block = { + .hw_ops = &vcn_v2_6_ras_hw_ops, + }, +}; + +static void vcn_v2_5_set_ras_funcs(struct amdgpu_device *adev) +{ + switch (adev->ip_versions[VCN_HWIP][0]) { + case IP_VERSION(2, 6, 0): + adev->vcn.ras = &vcn_v2_6_ras; + break; + default: + break; + } + + if (adev->vcn.ras) { + amdgpu_ras_register_ras_block(adev, &adev->vcn.ras->ras_block); + + strcpy(adev->vcn.ras->ras_block.ras_comm.name, "vcn"); + adev->vcn.ras->ras_block.ras_comm.block = AMDGPU_RAS_BLOCK__VCN; + adev->vcn.ras->ras_block.ras_comm.type = AMDGPU_RAS_ERROR__POISON; + adev->vcn.ras_if = &adev->vcn.ras->ras_block.ras_comm; + + /* If don't define special ras_late_init function, use default ras_late_init */ + if (!adev->vcn.ras->ras_block.ras_late_init) + adev->vcn.ras->ras_block.ras_late_init = amdgpu_ras_block_late_init; + } +} diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.h b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.h index e72f799ed0fd..1c19af74e4fd 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.h +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.h @@ -24,6 +24,12 @@ #ifndef __VCN_V2_5_H__ #define __VCN_V2_5_H__ +enum amdgpu_vcn_v2_6_sub_block { + AMDGPU_VCN_V2_6_VCPU_VCODEC = 0, + + AMDGPU_VCN_V2_6_MAX_SUB_BLOCK, +}; + extern const struct amdgpu_ip_block_version vcn_v2_5_ip_block; extern const struct amdgpu_ip_block_version vcn_v2_6_ip_block; -- cgit v1.2.3 From 749831acb1f6bf5837aeeec93b96da1d91534775 Mon Sep 17 00:00:00 2001 From: Mohammad Zafar Ziya Date: Wed, 23 Mar 2022 12:10:00 +0800 Subject: drm/amdgpu/jpeg: Add jpeg ras error query support RAS error query support addition for JPEG 2.6 V2: removed unused options and corrected comment format. Moved register definition to header file. V3: poison query status check added. Removed the error query support V4: Return statement refactored. Signed-off-by: Mohammad Zafar Ziya Reviewed-by: Hawking Zhang Reviewed-by: Tao Zhou Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c | 74 ++++++++++++++++++++++++++++++++++ drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.h | 7 ++++ 2 files changed, 81 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c index a29c86617fb5..8c3227d0b8b4 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.c @@ -26,6 +26,7 @@ #include "soc15.h" #include "soc15d.h" #include "jpeg_v2_0.h" +#include "jpeg_v2_5.h" #include "vcn/vcn_2_5_offset.h" #include "vcn/vcn_2_5_sh_mask.h" @@ -39,6 +40,7 @@ static void jpeg_v2_5_set_dec_ring_funcs(struct amdgpu_device *adev); static void jpeg_v2_5_set_irq_funcs(struct amdgpu_device *adev); static int jpeg_v2_5_set_powergating_state(void *handle, enum amd_powergating_state state); +static void jpeg_v2_5_set_ras_funcs(struct amdgpu_device *adev); static int amdgpu_ih_clientid_jpeg[] = { SOC15_IH_CLIENTID_VCN, @@ -70,6 +72,7 @@ static int jpeg_v2_5_early_init(void *handle) jpeg_v2_5_set_dec_ring_funcs(adev); jpeg_v2_5_set_irq_funcs(adev); + jpeg_v2_5_set_ras_funcs(adev); return 0; } @@ -730,3 +733,74 @@ const struct amdgpu_ip_block_version jpeg_v2_6_ip_block = .rev = 0, .funcs = &jpeg_v2_6_ip_funcs, }; + +static uint32_t jpeg_v2_6_query_poison_by_instance(struct amdgpu_device *adev, + uint32_t instance, uint32_t sub_block) +{ + uint32_t poison_stat = 0, reg_value = 0; + + switch (sub_block) { + case AMDGPU_JPEG_V2_6_JPEG0: + reg_value = RREG32_SOC15(JPEG, instance, mmUVD_RAS_JPEG0_STATUS); + poison_stat = REG_GET_FIELD(reg_value, UVD_RAS_JPEG0_STATUS, POISONED_PF); + break; + case AMDGPU_JPEG_V2_6_JPEG1: + reg_value = RREG32_SOC15(JPEG, instance, mmUVD_RAS_JPEG1_STATUS); + poison_stat = REG_GET_FIELD(reg_value, UVD_RAS_JPEG1_STATUS, POISONED_PF); + break; + default: + break; + } + + if (poison_stat) + dev_info(adev->dev, "Poison detected in JPEG%d sub_block%d\n", + instance, sub_block); + + return poison_stat; +} + +static bool jpeg_v2_6_query_ras_poison_status(struct amdgpu_device *adev) +{ + uint32_t inst = 0, sub = 0, poison_stat = 0; + + for (inst = 0; inst < adev->jpeg.num_jpeg_inst; inst++) + for (sub = 0; sub < AMDGPU_JPEG_V2_6_MAX_SUB_BLOCK; sub++) + poison_stat += + jpeg_v2_6_query_poison_by_instance(adev, inst, sub); + + return !!poison_stat; +} + +const struct amdgpu_ras_block_hw_ops jpeg_v2_6_ras_hw_ops = { + .query_poison_status = jpeg_v2_6_query_ras_poison_status, +}; + +static struct amdgpu_jpeg_ras jpeg_v2_6_ras = { + .ras_block = { + .hw_ops = &jpeg_v2_6_ras_hw_ops, + }, +}; + +static void jpeg_v2_5_set_ras_funcs(struct amdgpu_device *adev) +{ + switch (adev->ip_versions[JPEG_HWIP][0]) { + case IP_VERSION(2, 6, 0): + adev->jpeg.ras = &jpeg_v2_6_ras; + break; + default: + break; + } + + if (adev->jpeg.ras) { + amdgpu_ras_register_ras_block(adev, &adev->jpeg.ras->ras_block); + + strcpy(adev->jpeg.ras->ras_block.ras_comm.name, "jpeg"); + adev->jpeg.ras->ras_block.ras_comm.block = AMDGPU_RAS_BLOCK__JPEG; + adev->jpeg.ras->ras_block.ras_comm.type = AMDGPU_RAS_ERROR__POISON; + adev->jpeg.ras_if = &adev->jpeg.ras->ras_block.ras_comm; + + /* If don't define special ras_late_init function, use default ras_late_init */ + if (!adev->jpeg.ras->ras_block.ras_late_init) + adev->jpeg.ras->ras_block.ras_late_init = amdgpu_ras_block_late_init; + } +} diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.h b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.h index 3b0aa29b9879..1e858c6cdf13 100644 --- a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.h +++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_5.h @@ -24,6 +24,13 @@ #ifndef __JPEG_V2_5_H__ #define __JPEG_V2_5_H__ +enum amdgpu_jpeg_v2_6_sub_block { + AMDGPU_JPEG_V2_6_JPEG0 = 0, + AMDGPU_JPEG_V2_6_JPEG1, + + AMDGPU_JPEG_V2_6_MAX_SUB_BLOCK, +}; + extern const struct amdgpu_ip_block_version jpeg_v2_5_ip_block; extern const struct amdgpu_ip_block_version jpeg_v2_6_ip_block; -- cgit v1.2.3 From 1aa6b091f622bab0a805edc52bbb95cc96d55f30 Mon Sep 17 00:00:00 2001 From: Angus Wang Date: Tue, 8 Mar 2022 09:33:24 -0500 Subject: drm/amd/display: Create underflow interrupt IRQ type [WHY] We want another entry in IRQ type that can be used to help find the underflow interrupt source. [HOW] Added another mapping in IRQ type enum. Reviewed-by: Jun Lei Acked-by: Alex Hung Signed-off-by: Angus Wang Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/irq_types.h | 1 + 1 file changed, 1 insertion(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/irq_types.h b/drivers/gpu/drm/amd/display/dc/irq_types.h index 530c2578db40..36d8243cfbff 100644 --- a/drivers/gpu/drm/amd/display/dc/irq_types.h +++ b/drivers/gpu/drm/amd/display/dc/irq_types.h @@ -162,6 +162,7 @@ enum irq_type IRQ_TYPE_VUPDATE = DC_IRQ_SOURCE_VUPDATE1, IRQ_TYPE_VBLANK = DC_IRQ_SOURCE_VBLANK1, IRQ_TYPE_VLINE0 = DC_IRQ_SOURCE_DC1_VLINE0, + IRQ_TYPE_DCUNDERFLOW = DC_IRQ_SOURCE_DC1UNDERFLOW, }; #define DAL_VALID_IRQ_SRC_NUM(src) \ -- cgit v1.2.3 From 71d7a553645f9aba80b318ee5b4385d264ac3efa Mon Sep 17 00:00:00 2001 From: Oliver Logush Date: Thu, 10 Mar 2022 11:16:09 -0500 Subject: drm/amd/display: Remove SW w/a for HDCP 1.4 1A-07 failure based on ECO fix [why] W/a no longer needed Reviewed-by: Charlene Liu Acked-by: Alex Hung Signed-off-by: Oliver Logush Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c | 3 +-- drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c | 3 +-- 2 files changed, 2 insertions(+), 4 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c b/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c index fadb89326999..e6f9312e3a48 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c @@ -1758,11 +1758,10 @@ static bool dcn315_resource_construct( pool->base.mpcc_count = pool->base.res_cap->num_timing_generator; dc->caps.max_downscale_ratio = 600; dc->caps.i2c_speed_in_khz = 100; - dc->caps.i2c_speed_in_khz_hdcp = 5; /*1.4 w/a applied by default*/ + dc->caps.i2c_speed_in_khz_hdcp = 100; dc->caps.max_cursor_size = 256; dc->caps.min_horizontal_blanking_period = 80; dc->caps.dmdata_alloc_size = 2048; - dc->caps.max_slave_planes = 1; dc->caps.max_slave_yuv_planes = 1; dc->caps.max_slave_rgb_planes = 1; diff --git a/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c b/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c index d73145dab173..d5c195749a81 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn316/dcn316_resource.c @@ -1760,11 +1760,10 @@ static bool dcn316_resource_construct( pool->base.mpcc_count = pool->base.res_cap->num_timing_generator; dc->caps.max_downscale_ratio = 600; dc->caps.i2c_speed_in_khz = 100; - dc->caps.i2c_speed_in_khz_hdcp = 5; /*1.4 w/a applied by default*/ + dc->caps.i2c_speed_in_khz_hdcp = 100; dc->caps.max_cursor_size = 256; dc->caps.min_horizontal_blanking_period = 80; dc->caps.dmdata_alloc_size = 2048; - dc->caps.max_slave_planes = 1; dc->caps.max_slave_yuv_planes = 1; dc->caps.max_slave_rgb_planes = 1; -- cgit v1.2.3 From 5abef8e5ce910545079f078ba4383f2650389e00 Mon Sep 17 00:00:00 2001 From: Jingwen Zhu Date: Tue, 15 Mar 2022 13:15:27 +0800 Subject: drm/amd/display: Set fec register init value [Why] We don't include this eDP FEC init on fastboot. [How] Set the fec to init value when stopping driver &get the fec register value to check should enable FEC. Co-authored-by: Jingwen Zhu Reviewed-by: Wenjing Liu Acked-by: Alex Hung Signed-off-by: Jingwen Zhu Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc_link.c | 26 ++++++++++------------ .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 6 ++++- drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c | 6 ++++- 3 files changed, 22 insertions(+), 16 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index bbaa5abdf888..c7c4d9867c52 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -4683,22 +4683,20 @@ bool dc_link_is_fec_supported(const struct dc_link *link) bool dc_link_should_enable_fec(const struct dc_link *link) { - bool is_fec_disable = false; - bool ret = false; + bool force_disable = false; - if ((link->connector_signal != SIGNAL_TYPE_DISPLAY_PORT_MST && + if (link->fec_state == dc_link_fec_enabled) + force_disable = false; + else if (link->connector_signal != SIGNAL_TYPE_DISPLAY_PORT_MST && link->local_sink && - link->local_sink->edid_caps.panel_patch.disable_fec) || - (link->connector_signal == SIGNAL_TYPE_EDP - // enable FEC for EDP if DSC is supported - && link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_SUPPORT == false - )) - is_fec_disable = true; - - if (dc_link_is_fec_supported(link) && !link->dc->debug.disable_fec && !is_fec_disable) - ret = true; - - return ret; + link->local_sink->edid_caps.panel_patch.disable_fec) + force_disable = true; + else if (link->connector_signal == SIGNAL_TYPE_EDP + && link->dpcd_caps.dsc_caps.dsc_basic_caps.fields. + dsc_support.DSC_SUPPORT == false) + force_disable = true; + + return !force_disable && dc_link_is_fec_supported(link); } uint32_t dc_bandwidth_in_kbps_from_timing( diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index ad757b59e00e..911c5d103c64 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c @@ -1493,8 +1493,12 @@ void dcn10_init_hw(struct dc *dc) /* Check for enabled DIG to identify enabled display */ if (link->link_enc->funcs->is_dig_enabled && - link->link_enc->funcs->is_dig_enabled(link->link_enc)) + link->link_enc->funcs->is_dig_enabled(link->link_enc)) { link->link_status.link_active = true; + if (link->link_enc->funcs->fec_is_active && + link->link_enc->funcs->fec_is_active(link->link_enc)) + link->fec_state = dc_link_fec_enabled; + } } /* Power gate DSCs */ diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c index 4be228680909..b57f657c4e44 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c @@ -188,8 +188,12 @@ void dcn31_init_hw(struct dc *dc) /* Check for enabled DIG to identify enabled display */ if (link->link_enc->funcs->is_dig_enabled && - link->link_enc->funcs->is_dig_enabled(link->link_enc)) + link->link_enc->funcs->is_dig_enabled(link->link_enc)) { link->link_status.link_active = true; + if (link->link_enc->funcs->fec_is_active && + link->link_enc->funcs->fec_is_active(link->link_enc)) + link->fec_state = dc_link_fec_enabled; + } } /* Enables outbox notifications for usb4 dpia */ -- cgit v1.2.3 From e9ebc23b3fb2949e18c6df98a6e6fd94429e498e Mon Sep 17 00:00:00 2001 From: Krunoslav Kovac Date: Fri, 11 Mar 2022 16:12:58 -0500 Subject: drm/amd/display: Enable 3-plane MPO for DCN31 [WHY&HOW] It can be enabled by users, but proper way is to report max_slave_planes in DC caps for each ASIC. Some structures use hardcoded max_plane=2, this is also addressed here. Reviewed-by: Nevenko Stupar Reviewed-by: Aric Cyr Acked-by: Alex Hung Signed-off-by: Krunoslav Kovac Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c index bf130b2435ab..826970f2bd0a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c @@ -1868,9 +1868,9 @@ static bool dcn31_resource_construct( dc->caps.min_horizontal_blanking_period = 80; dc->caps.dmdata_alloc_size = 2048; - dc->caps.max_slave_planes = 1; - dc->caps.max_slave_yuv_planes = 1; - dc->caps.max_slave_rgb_planes = 1; + dc->caps.max_slave_planes = 2; + dc->caps.max_slave_yuv_planes = 2; + dc->caps.max_slave_rgb_planes = 2; dc->caps.post_blend_color_processing = true; dc->caps.force_dp_tps4_for_cp2520 = true; dc->caps.dp_hpo = true; -- cgit v1.2.3 From 83bb503275bdf651c67f02e0d25f2d0db2ca865b Mon Sep 17 00:00:00 2001 From: Chris Park Date: Tue, 15 Mar 2022 12:21:43 -0400 Subject: drm/amd/display: Correct Slice reset calculation [Why] Once DSC slice cannot fit pixel clock, we incorrectly reset min slices to 0 and allow max slice to operate, even when max slice itself cannot fit the pixel clock properly. [How] Change the sequence such that we correctly determine DSC is not possible when both min slices and max slices cannot fit pixel clock per slice. Reviewed-by: Wenjing Liu Acked-by: Alex Hung Signed-off-by: Chris Park Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c b/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c index efc2339f1fa0..4385d19bc489 100644 --- a/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c +++ b/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c @@ -864,11 +864,11 @@ static bool setup_dsc_config( min_slices_h = inc_num_slices(dsc_common_caps.slice_caps, min_slices_h); } + is_dsc_possible = (min_slices_h <= max_slices_h); + if (pic_width % min_slices_h != 0) min_slices_h = 0; // DSC TODO: Maybe try increasing the number of slices first? - is_dsc_possible = (min_slices_h <= max_slices_h); - if (min_slices_h == 0 && max_slices_h == 0) is_dsc_possible = false; -- cgit v1.2.3 From 06bab5897c51e15d9a6b7ec11b1eacc4f2e1a145 Mon Sep 17 00:00:00 2001 From: Paul Hsieh Date: Wed, 16 Mar 2022 06:59:46 +0800 Subject: drm/amd/display: Power down hardware if timer not trigger [WHY] In headless systems, if SetMode/Power down timer is not called, hardware will not be powered down causing HW/SW discrepancies. Powering down hardware on SetPowerState to D3 will ensure SW/HW state is accurate. [HOW] If PowerDownThread timer is not trigger but OS call SetPowerState to D3, power down hardware. Reviewed-by: Eric Yang Acked-by: Alex Hung Signed-off-by: Paul Hsieh Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- .../amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c | 26 +++++++++++++++++++++- 1 file changed, 25 insertions(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c index 59fdd7f0d609..969b40250434 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c @@ -615,13 +615,37 @@ static void dcn31_clk_mgr_helper_populate_bw_params(struct clk_mgr_internal *clk } } +void dcn31_set_low_power_state(struct clk_mgr *clk_mgr_base) +{ + int display_count; + struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); + struct dc *dc = clk_mgr_base->ctx->dc; + struct dc_state *context = dc->current_state; + + if (clk_mgr_base->clks.pwr_state != DCN_PWR_STATE_LOW_POWER) { + display_count = dcn31_get_active_display_cnt_wa(dc, context); + /* if we can go lower, go lower */ + if (display_count == 0) { + union display_idle_optimization_u idle_info = { 0 }; + + idle_info.idle_info.df_request_disabled = 1; + idle_info.idle_info.phy_ref_clk_off = 1; + idle_info.idle_info.s0i2_rdy = 1; + dcn31_smu_set_display_idle_optimization(clk_mgr, idle_info.data); + /* update power state */ + clk_mgr_base->clks.pwr_state = DCN_PWR_STATE_LOW_POWER; + } + } +} + static struct clk_mgr_funcs dcn31_funcs = { .get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz, .update_clocks = dcn31_update_clocks, .init_clocks = dcn31_init_clocks, .enable_pme_wa = dcn31_enable_pme_wa, .are_clock_states_equal = dcn31_are_clock_states_equal, - .notify_wm_ranges = dcn31_notify_wm_ranges + .notify_wm_ranges = dcn31_notify_wm_ranges, + .set_low_power_state = dcn31_set_low_power_state }; extern struct clk_mgr_funcs dcn3_fpga_funcs; -- cgit v1.2.3 From 75c95f4550e3d4584ff100fef53c240f3291f7cd Mon Sep 17 00:00:00 2001 From: Roman Li Date: Tue, 15 Mar 2022 14:57:34 -0400 Subject: drm/amd/display: Remove redundant dsc power gating from init_hw [Why] DSC Power down code has been moved from dcn31_init_hw into init_pipes() Need to remove it from dcn10_init_hw() as well to avoid duplicated action on dcn1.x/2.x [How] Remove DSC power down code from dcn10_init_hw() Fixes: 8fa6f4c5715c ("drm/amd/display: fixed the DSC power off sequence during Driver PnP") Reviewed-by: Anthony Koo Reviewed-by: Eric Yang Acked-by: Alex Hung Signed-off-by: Roman Li Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 7 ------- 1 file changed, 7 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index 911c5d103c64..bb309ccee3e4 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c @@ -1501,13 +1501,6 @@ void dcn10_init_hw(struct dc *dc) } } - /* Power gate DSCs */ - if (!is_optimized_init_done) { - for (i = 0; i < res_pool->res_cap->num_dsc; i++) - if (hws->funcs.dsc_pg_control != NULL) - hws->funcs.dsc_pg_control(hws, res_pool->dscs[i]->inst, false); - } - /* we want to turn off all dp displays before doing detection */ dc_link_blank_all_dp_displays(dc); -- cgit v1.2.3 From 384bd90d1cc1168dc845ad2735d7a3d49459fbba Mon Sep 17 00:00:00 2001 From: Roman Li Date: Tue, 15 Mar 2022 16:31:14 -0400 Subject: drm/amd/display: Enable power gating before init_pipes [Why] In init_hw() we call init_pipes() before enabling power gating. init_pipes() tries to power gate dsc but it may fail because required force-ons are not released yet. As a result with dsc config the following errors observed on resume: "REG_WAIT timeout 1us * 1000 tries - dcn20_dsc_pg_control" "REG_WAIT timeout 1us * 1000 tries - dcn20_dpp_pg_control" "REG_WAIT timeout 1us * 1000 tries - dcn20_hubp_pg_control" [How] Move enable_power_gating_plane() before init_pipes() in init_hw() Reviewed-by: Anthony Koo Reviewed-by: Eric Yang Acked-by: Alex Hung Signed-off-by: Roman Li Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 5 +++-- drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c | 5 +++-- drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c | 5 +++-- 3 files changed, 9 insertions(+), 6 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index bb309ccee3e4..e4247740ac12 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c @@ -1504,6 +1504,9 @@ void dcn10_init_hw(struct dc *dc) /* we want to turn off all dp displays before doing detection */ dc_link_blank_all_dp_displays(dc); + if (hws->funcs.enable_power_gating_plane) + hws->funcs.enable_power_gating_plane(dc->hwseq, true); + /* If taking control over from VBIOS, we may want to optimize our first * mode set, so we need to skip powering down pipes until we know which * pipes we want to use. @@ -1556,8 +1559,6 @@ void dcn10_init_hw(struct dc *dc) REG_UPDATE(DCFCLK_CNTL, DCFCLK_GATE_DIS, 0); } - if (hws->funcs.enable_power_gating_plane) - hws->funcs.enable_power_gating_plane(dc->hwseq, true); if (dc->clk_mgr->funcs->notify_wm_ranges) dc->clk_mgr->funcs->notify_wm_ranges(dc->clk_mgr); diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c index ed0a0e5fd805..f61ec8763844 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c @@ -547,6 +547,9 @@ void dcn30_init_hw(struct dc *dc) /* we want to turn off all dp displays before doing detection */ dc_link_blank_all_dp_displays(dc); + if (hws->funcs.enable_power_gating_plane) + hws->funcs.enable_power_gating_plane(dc->hwseq, true); + /* If taking control over from VBIOS, we may want to optimize our first * mode set, so we need to skip powering down pipes until we know which * pipes we want to use. @@ -624,8 +627,6 @@ void dcn30_init_hw(struct dc *dc) REG_UPDATE(DCFCLK_CNTL, DCFCLK_GATE_DIS, 0); } - if (hws->funcs.enable_power_gating_plane) - hws->funcs.enable_power_gating_plane(dc->hwseq, true); if (!dcb->funcs->is_accelerated_mode(dcb) && dc->res_pool->hubbub->funcs->init_watermarks) dc->res_pool->hubbub->funcs->init_watermarks(dc->res_pool->hubbub); diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c index b57f657c4e44..67c13654ab99 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c @@ -203,6 +203,9 @@ void dcn31_init_hw(struct dc *dc) /* we want to turn off all dp displays before doing detection */ dc_link_blank_all_dp_displays(dc); + if (hws->funcs.enable_power_gating_plane) + hws->funcs.enable_power_gating_plane(dc->hwseq, true); + /* If taking control over from VBIOS, we may want to optimize our first * mode set, so we need to skip powering down pipes until we know which * pipes we want to use. @@ -252,8 +255,6 @@ void dcn31_init_hw(struct dc *dc) REG_UPDATE(DCFCLK_CNTL, DCFCLK_GATE_DIS, 0); } - if (hws->funcs.enable_power_gating_plane) - hws->funcs.enable_power_gating_plane(dc->hwseq, true); if (!dcb->funcs->is_accelerated_mode(dcb) && dc->res_pool->hubbub->funcs->init_watermarks) dc->res_pool->hubbub->funcs->init_watermarks(dc->res_pool->hubbub); -- cgit v1.2.3 From a905f0b59fefe819c59bacf9d0cfa678c96fcb8c Mon Sep 17 00:00:00 2001 From: Charlene Liu Date: Tue, 15 Mar 2022 16:42:33 -0400 Subject: drm/amd/display: Clear optc false state when disable otg [why] when disable optc, need to clear the underflow status as well. Reviewed-by: Chris Park Acked-by: Alex Hung Signed-off-by: Charlene Liu Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c | 14 +++++++------- drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c | 5 ++++- 2 files changed, 11 insertions(+), 8 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c index 67c13654ab99..531dd2c65007 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c @@ -343,20 +343,20 @@ void dcn31_enable_power_gating_plane( bool enable) { bool force_on = true; /* disable power gating */ + uint32_t org_ip_request_cntl = 0; if (enable && !hws->ctx->dc->debug.disable_hubp_power_gate) force_on = false; + REG_GET(DC_IP_REQUEST_CNTL, IP_REQUEST_EN, &org_ip_request_cntl); + if (org_ip_request_cntl == 0) + REG_SET(DC_IP_REQUEST_CNTL, 0, IP_REQUEST_EN, 1); /* DCHUBP0/1/2/3/4/5 */ REG_UPDATE(DOMAIN0_PG_CONFIG, DOMAIN_POWER_FORCEON, force_on); - REG_WAIT(DOMAIN0_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, force_on, 1, 1000); REG_UPDATE(DOMAIN2_PG_CONFIG, DOMAIN_POWER_FORCEON, force_on); - REG_WAIT(DOMAIN2_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, force_on, 1, 1000); /* DPP0/1/2/3/4/5 */ REG_UPDATE(DOMAIN1_PG_CONFIG, DOMAIN_POWER_FORCEON, force_on); - REG_WAIT(DOMAIN1_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, force_on, 1, 1000); REG_UPDATE(DOMAIN3_PG_CONFIG, DOMAIN_POWER_FORCEON, force_on); - REG_WAIT(DOMAIN3_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, force_on, 1, 1000); force_on = true; /* disable power gating */ if (enable && !hws->ctx->dc->debug.disable_dsc_power_gate) @@ -364,11 +364,11 @@ void dcn31_enable_power_gating_plane( /* DCS0/1/2/3/4/5 */ REG_UPDATE(DOMAIN16_PG_CONFIG, DOMAIN_POWER_FORCEON, force_on); - REG_WAIT(DOMAIN16_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, force_on, 1, 1000); REG_UPDATE(DOMAIN17_PG_CONFIG, DOMAIN_POWER_FORCEON, force_on); - REG_WAIT(DOMAIN17_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, force_on, 1, 1000); REG_UPDATE(DOMAIN18_PG_CONFIG, DOMAIN_POWER_FORCEON, force_on); - REG_WAIT(DOMAIN18_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, force_on, 1, 1000); + + if (org_ip_request_cntl == 0) + REG_SET(DC_IP_REQUEST_CNTL, 0, IP_REQUEST_EN, 0); } void dcn31_update_info_frame(struct pipe_ctx *pipe_ctx) diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c index 8afe2130d7c5..e05527a3a8ba 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c @@ -124,7 +124,6 @@ static bool optc31_enable_crtc(struct timing_generator *optc) static bool optc31_disable_crtc(struct timing_generator *optc) { struct optc *optc1 = DCN10TG_FROM_TG(optc); - /* disable otg request until end of the first line * in the vertical blank region */ @@ -138,6 +137,7 @@ static bool optc31_disable_crtc(struct timing_generator *optc) REG_WAIT(OTG_CLOCK_CONTROL, OTG_BUSY, 0, 1, 100000); + optc1_clear_optc_underflow(optc); return true; } @@ -158,6 +158,9 @@ static bool optc31_immediate_disable_crtc(struct timing_generator *optc) OTG_BUSY, 0, 1, 100000); + /* clear the false state */ + optc1_clear_optc_underflow(optc); + return true; } -- cgit v1.2.3 From 84d1ce0631c746c7fe025cbb49da6ade1aa80a1f Mon Sep 17 00:00:00 2001 From: Jimmy Kizito Date: Wed, 9 Mar 2022 18:26:47 -0500 Subject: drm/amd/display: Add work around for AUX failure on wake. [Why] When waking from low-power states, a DP sink may remain unresponsive to AUX transactions. [How] Try to toggle DPCD SET_POWER register repeatedly (up to a maximum timeout value) until DP sink becomes responsive. Reviewed-by: Mustapha Ghaddar Acked-by: Alex Hung Signed-off-by: Jimmy Kizito Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c | 59 ++++++++++++++++++++++++ drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h | 1 + 2 files changed, 60 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c index 351081f574cb..e4df81dc1dc2 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c @@ -5216,6 +5216,62 @@ static void retrieve_cable_id(struct dc_link *link) &link->dpcd_caps.cable_id, &usbc_cable_id); } +/* DPRX may take some time to respond to AUX messages after HPD asserted. + * If AUX read unsuccessful, try to wake unresponsive DPRX by toggling DPCD SET_POWER (0x600). + */ +static enum dc_status wa_try_to_wake_dprx(struct dc_link *link, uint64_t timeout_ms) +{ + enum dc_status status = DC_ERROR_UNEXPECTED; + uint8_t dpcd_data = 0; + uint64_t start_ts = 0; + uint64_t current_ts = 0; + uint64_t time_taken_ms = 0; + enum dc_connection_type type = dc_connection_none; + + status = core_link_read_dpcd( + link, + DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV, + &dpcd_data, + sizeof(dpcd_data)); + + if (status != DC_OK) { + DC_LOG_WARNING("%s: Read DPCD LTTPR_CAP failed - try to toggle DPCD SET_POWER for %lld ms.", + __func__, + timeout_ms); + start_ts = dm_get_timestamp(link->ctx); + + do { + if (!dc_link_detect_sink(link, &type) || type == dc_connection_none) + break; + + dpcd_data = DP_SET_POWER_D3; + status = core_link_write_dpcd( + link, + DP_SET_POWER, + &dpcd_data, + sizeof(dpcd_data)); + + dpcd_data = DP_SET_POWER_D0; + status = core_link_write_dpcd( + link, + DP_SET_POWER, + &dpcd_data, + sizeof(dpcd_data)); + + current_ts = dm_get_timestamp(link->ctx); + time_taken_ms = div_u64(dm_get_elapse_time_in_ns(link->ctx, current_ts, start_ts), 1000000); + } while (status != DC_OK && time_taken_ms < timeout_ms); + + DC_LOG_WARNING("%s: DPCD SET_POWER %s after %lld ms%s", + __func__, + (status == DC_OK) ? "succeeded" : "failed", + time_taken_ms, + (type == dc_connection_none) ? ". Unplugged." : "."); + } + + return status; +} + static bool retrieve_link_cap(struct dc_link *link) { /* DP_ADAPTER_CAP - DP_DPCD_REV + 1 == 16 and also DP_DSC_BITS_PER_PIXEL_INC - DP_DSC_SUPPORT + 1 == 16, @@ -5251,6 +5307,9 @@ static bool retrieve_link_cap(struct dc_link *link) dc_link_aux_try_to_configure_timeout(link->ddc, LINK_AUX_DEFAULT_LTTPR_TIMEOUT_PERIOD); + /* Try to ensure AUX channel active before proceeding. */ + status = wa_try_to_wake_dprx(link, LINK_AUX_WAKE_TIMEOUT_MS); + is_lttpr_present = dp_retrieve_lttpr_cap(link); /* Read DP tunneling information. */ status = dpcd_get_tunneling_device_data(link); diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h index ab9939db8cea..44f167d2584f 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h @@ -33,6 +33,7 @@ #define MAX_MTP_SLOT_COUNT 64 #define DP_REPEATER_CONFIGURATION_AND_STATUS_SIZE 0x50 #define TRAINING_AUX_RD_INTERVAL 100 //us +#define LINK_AUX_WAKE_TIMEOUT_MS 1500 // Timeout when trying to wake unresponsive DPRX. struct dc_link; struct dc_stream_state; -- cgit v1.2.3 From 8cb3c169fc7574cb889deb3b59a3c8e783fb4685 Mon Sep 17 00:00:00 2001 From: Samson Tam Date: Wed, 16 Mar 2022 01:06:42 -0400 Subject: drm/amd/display: Add support for USBC connector [Why] Add support for CONNECTOR_ID_USBC Reviewed-by: Alvin Lee Acked-by: Alex Hung Signed-off-by: Samson Tam Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/bios/bios_parser_common.c | 3 +++ drivers/gpu/drm/amd/display/dc/bios/command_table.c | 3 ++- drivers/gpu/drm/amd/display/dc/core/dc_link.c | 8 ++++++-- drivers/gpu/drm/amd/display/include/grph_object_id.h | 1 + 4 files changed, 12 insertions(+), 3 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser_common.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser_common.c index a8cb039d2572..34e3a64f556e 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser_common.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser_common.c @@ -213,6 +213,9 @@ static enum connector_id connector_id_from_bios_object_id( case CONNECTOR_OBJECT_ID_MXM: id = CONNECTOR_ID_MXM; break; + case CONNECTOR_OBJECT_ID_USBC: + id = CONNECTOR_ID_USBC; + break; default: id = CONNECTOR_ID_UNKNOWN; break; diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table.c b/drivers/gpu/drm/amd/display/dc/bios/command_table.c index 0e36cd800fc9..32efa92422e8 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/command_table.c +++ b/drivers/gpu/drm/amd/display/dc/bios/command_table.c @@ -522,7 +522,8 @@ static enum bp_result transmitter_control_v2( */ params.acConfig.ucEncoderSel = 1; - if (CONNECTOR_ID_DISPLAY_PORT == connector_id) + if (CONNECTOR_ID_DISPLAY_PORT == connector_id + || CONNECTOR_ID_USBC == connector_id) /* Bit4: DP connector flag * =0 connector is none-DP connector * =1 connector is DP connector diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index c7c4d9867c52..7aede6495e5e 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -345,6 +345,7 @@ static enum signal_type get_basic_signal_type(struct graphics_object_id encoder, case CONNECTOR_ID_LVDS: return SIGNAL_TYPE_LVDS; case CONNECTOR_ID_DISPLAY_PORT: + case CONNECTOR_ID_USBC: return SIGNAL_TYPE_DISPLAY_PORT; case CONNECTOR_ID_EDP: return SIGNAL_TYPE_EDP; @@ -380,7 +381,8 @@ bool dc_link_is_dp_sink_present(struct dc_link *link) bool present = ((connector_id == CONNECTOR_ID_DISPLAY_PORT) || - (connector_id == CONNECTOR_ID_EDP)); + (connector_id == CONNECTOR_ID_EDP) || + (connector_id == CONNECTOR_ID_USBC)); ddc = dal_ddc_service_get_ddc_pin(link->ddc); @@ -476,7 +478,8 @@ static enum signal_type link_detect_sink(struct dc_link *link, result = SIGNAL_TYPE_DVI_SINGLE_LINK; } break; - case CONNECTOR_ID_DISPLAY_PORT: { + case CONNECTOR_ID_DISPLAY_PORT: + case CONNECTOR_ID_USBC: { /* DP HPD short pulse. Passive DP dongle will not * have short pulse */ @@ -1591,6 +1594,7 @@ static bool dc_link_construct_legacy(struct dc_link *link, link->connector_signal = SIGNAL_TYPE_DVI_DUAL_LINK; break; case CONNECTOR_ID_DISPLAY_PORT: + case CONNECTOR_ID_USBC: link->connector_signal = SIGNAL_TYPE_DISPLAY_PORT; if (link->hpd_gpio) diff --git a/drivers/gpu/drm/amd/display/include/grph_object_id.h b/drivers/gpu/drm/amd/display/include/grph_object_id.h index fed1edc038d8..c6bbd262f1ac 100644 --- a/drivers/gpu/drm/amd/display/include/grph_object_id.h +++ b/drivers/gpu/drm/amd/display/include/grph_object_id.h @@ -162,6 +162,7 @@ enum connector_id { CONNECTOR_ID_MXM = 21, CONNECTOR_ID_WIRELESS = 22, CONNECTOR_ID_MIRACAST = 23, + CONNECTOR_ID_USBC = 24, CONNECTOR_ID_VIRTUAL = 100 }; -- cgit v1.2.3 From a772105cfaca3c7b28bbe8a4e22a08efa561b603 Mon Sep 17 00:00:00 2001 From: Michael Strauss Date: Thu, 17 Mar 2022 10:17:01 -0400 Subject: drm/amd/display: Update LTTPR UHBR link rate support struct [WHY] Update field order to match DP2.0 spec SCR Reviewed-by: George Shen Acked-by: Alex Hung Signed-off-by: Michael Strauss Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dc_dp_types.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h index 36ac2a8746bd..7d4aa99525da 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h @@ -993,8 +993,8 @@ union dp_128b_132b_supported_link_rates { union dp_128b_132b_supported_lttpr_link_rates { struct { uint8_t UHBR10 :1; - uint8_t UHBR13_5:1; uint8_t UHBR20 :1; + uint8_t UHBR13_5:1; uint8_t RESERVED:5; } bits; uint8_t raw; -- cgit v1.2.3 From 4304b8a434b6b95449cd81f02e953f9de357a707 Mon Sep 17 00:00:00 2001 From: Martin Leung Date: Fri, 18 Mar 2022 11:12:36 -0400 Subject: drm/amd/display: Revert FEC check in validation why and how: causes failure on install on certain machines Reviewed-by: George Shen Acked-by: Alex Hung Signed-off-by: Martin Leung Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc.c | 4 ---- 1 file changed, 4 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index f2ad8f58e69c..c436db416708 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -1496,10 +1496,6 @@ bool dc_validate_boot_timing(const struct dc *dc, if (!link->link_enc->funcs->is_dig_enabled(link->link_enc)) return false; - /* Check for FEC status*/ - if (link->link_enc->funcs->fec_is_active(link->link_enc)) - return false; - enc_inst = link->link_enc->funcs->get_dig_frontend(link->link_enc); if (enc_inst == ENGINE_ID_UNKNOWN) -- cgit v1.2.3 From ad371c8a2f66c5e14a8baa26472ec833ffa8342b Mon Sep 17 00:00:00 2001 From: Anthony Koo Date: Sat, 19 Mar 2022 13:09:57 -0400 Subject: drm/amd/display: [FW Promotion] Release 0.0.110.0 - Revert save/restore PANEL_PWRSEQ_REF_DIV2 and other psr phy optimizations Reviewed-by: Aric Cyr Acked-by: Alex Hung Signed-off-by: Anthony Koo Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h | 90 ++----------------------- 1 file changed, 4 insertions(+), 86 deletions(-) (limited to 'drivers/gpu') 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 ce773b56a778..9b5db16b2619 100644 --- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h @@ -46,10 +46,10 @@ /* Firmware versioning. */ #ifdef DMUB_EXPOSE_VERSION -#define DMUB_FW_VERSION_GIT_HASH 0x51b95a35 +#define DMUB_FW_VERSION_GIT_HASH 0x19edd13d #define DMUB_FW_VERSION_MAJOR 0 #define DMUB_FW_VERSION_MINOR 0 -#define DMUB_FW_VERSION_REVISION 109 +#define DMUB_FW_VERSION_REVISION 110 #define DMUB_FW_VERSION_TEST 0 #define DMUB_FW_VERSION_VBIOS 0 #define DMUB_FW_VERSION_HOTFIX 0 @@ -1450,81 +1450,6 @@ enum dmub_cmd_mall_type { DMUB_CMD__MALL_ACTION_NO_DF_REQ = 3, }; -/** - * PHY Link rate for DP. - */ -enum phy_link_rate { - /** - * not supported. - */ - PHY_RATE_UNKNOWN = 0, - /** - * Rate_1 (RBR) - 1.62 Gbps/Lane - */ - PHY_RATE_162 = 1, - /** - * Rate_2 - 2.16 Gbps/Lane - */ - PHY_RATE_216 = 2, - /** - * Rate_3 - 2.43 Gbps/Lane - */ - PHY_RATE_243 = 3, - /** - * Rate_4 (HBR) - 2.70 Gbps/Lane - */ - PHY_RATE_270 = 4, - /** - * Rate_5 (RBR2)- 3.24 Gbps/Lane - */ - PHY_RATE_324 = 5, - /** - * Rate_6 - 4.32 Gbps/Lane - */ - PHY_RATE_432 = 6, - /** - * Rate_7 (HBR2)- 5.40 Gbps/Lane - */ - PHY_RATE_540 = 7, - /** - * Rate_8 (HBR3)- 8.10 Gbps/Lane - */ - PHY_RATE_810 = 8, - /** - * UHBR10 - 10.0 Gbps/Lane - */ - PHY_RATE_1000 = 9, - /** - * UHBR13.5 - 13.5 Gbps/Lane - */ - PHY_RATE_1350 = 10, - /** - * UHBR10 - 20.0 Gbps/Lane - */ - PHY_RATE_2000 = 11, -}; - -/** - * enum dmub_phy_fsm_state - PHY FSM states. - * PHY FSM state to transit to during PSR enable/disable. - */ -enum dmub_phy_fsm_state { - DMUB_PHY_FSM_POWER_UP_DEFAULT = 0, - DMUB_PHY_FSM_RESET, - DMUB_PHY_FSM_RESET_RELEASED, - DMUB_PHY_FSM_SRAM_LOAD_DONE, - DMUB_PHY_FSM_INITIALIZED, - DMUB_PHY_FSM_CALIBRATED, - DMUB_PHY_FSM_CALIBRATED_LP, - DMUB_PHY_FSM_CALIBRATED_PG, - DMUB_PHY_FSM_POWER_DOWN, - DMUB_PHY_FSM_PLL_EN, - DMUB_PHY_FSM_TX_EN, - DMUB_PHY_FSM_FAST_LP, -}; - - - /** * Data passed from driver to FW in a DMUB_CMD__PSR_COPY_SETTINGS command. */ @@ -1772,16 +1697,9 @@ struct dmub_cmd_psr_force_static_data { */ uint8_t panel_inst; /** - * Phy state to enter. - * Values to use are defined in dmub_phy_fsm_state - */ - uint8_t phy_fsm_state; - /** - * Phy rate for DP - RBR/HBR/HBR2/HBR3. - * Set this using enum phy_link_rate. - * This does not support HDMI/DP2 for now. + * Explicit padding to 4 byte boundary. */ - uint8_t phy_rate; + uint8_t pad[2]; }; /** -- cgit v1.2.3 From 9510d365e0626ff6f7ed5b23f70ae49828b68039 Mon Sep 17 00:00:00 2001 From: Aric Cyr Date: Sun, 20 Mar 2022 14:35:12 -0400 Subject: drm/amd/display: 3.2.179 - [FW Promotion] Release 0.0.110.0 - Revert FEC check in validation - Update LTTPR UHBR link rate support struct - Add support for USBC connector - Add work around for AUX failure on wake - Clear optc false state when disable otg - Enable power gating before init_pipes - Remove redundant dsc power gating from init_hw - Power down hardware if timer not trigger - Correct Slice reset calculation - Enable 3-plane MPO for DCN31 - Set fec register init value - Remove SW w/a for HDCP 1.4 1A-07 failure based on ECO fix - Create underflow interrupt IRQ type Acked-by: Alex Hung Signed-off-by: Aric Cyr 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(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index ced40fe218ac..ef286aa30294 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -47,7 +47,7 @@ struct aux_payload; struct set_config_cmd_payload; struct dmub_notification; -#define DC_VER "3.2.178" +#define DC_VER "3.2.179" #define MAX_SURFACES 3 #define MAX_PLANES 6 -- cgit v1.2.3 From c4af42b4eb5beb6f5ae96b5be786862bbb9cfb16 Mon Sep 17 00:00:00 2001 From: Roman Li Date: Thu, 17 Mar 2022 19:55:05 -0400 Subject: drm/amd/display: Fix allocate_mst_payload assert on resume [Why] On resume we do link detection for all non-MST connectors. MST is handled separately. However the condition for telling if connector is on mst branch is not enough for mst hub case. Link detection for mst branch link leads to mst topology reset. That causes assert in dc_link_allocate_mst_payload() [How] Use link type as indicator for mst link. Reviewed-by: Wayne Lin Acked-by: Alex Hung Signed-off-by: Roman Li Tested-by: Daniel Wheeler Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'drivers/gpu') 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 45d130f86114..35e714d28fee 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -2714,7 +2714,8 @@ static int dm_resume(void *handle) * this is the case when traversing through already created * MST connectors, should be skipped */ - if (aconnector->mst_port) + if (aconnector->dc_link && + aconnector->dc_link->type == dc_connection_mst_branch) continue; mutex_lock(&aconnector->hpd_lock); -- cgit v1.2.3 From 3222a81148ca273af66cb67f8445114bc633e747 Mon Sep 17 00:00:00 2001 From: Haowen Bai Date: Thu, 24 Mar 2022 17:27:14 +0800 Subject: drm/amd/display: Fix pointer dereferenced before checking The pointer edid_buf is dereferencing pointer edid before edid is being null checked. Fix this by assigning edid->raw_edid to edid_buf only if edid is not NULL, otherwise just NULL. Signed-off-by: Haowen Bai Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c index f5f39984702f..28cf24f6ab32 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c @@ -90,7 +90,7 @@ enum dc_edid_status dm_helpers_parse_edid_caps( { struct amdgpu_dm_connector *aconnector = link->priv; struct drm_connector *connector = &aconnector->base; - struct edid *edid_buf = (struct edid *) edid->raw_edid; + struct edid *edid_buf = edid ? (struct edid *) edid->raw_edid : NULL; struct cea_sad *sads; int sad_count = -1; int sadb_count = -1; -- cgit v1.2.3 From 9563e1ec9292254a5feb3dd1b7fbef0bed512c8f Mon Sep 17 00:00:00 2001 From: Philip Yang Date: Sun, 27 Mar 2022 22:02:56 -0400 Subject: drm/amdgpu: Correct unlocked update fence handling MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit To fix two issues with unlocked update fence: 1. vm->last_unlocked store the latest fence without taking refcount. 2. amdgpu_vm_bo_update_mapping returns old fence, not the latest fence. Signed-off-by: Philip Yang Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c index dbb551762805..69fba68ff88e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c @@ -109,7 +109,7 @@ static int amdgpu_vm_sdma_commit(struct amdgpu_vm_update_params *p, if (p->unlocked) { struct dma_fence *tmp = dma_fence_get(f); - swap(p->vm->last_unlocked, f); + swap(p->vm->last_unlocked, tmp); dma_fence_put(tmp); } else { amdgpu_bo_fence(p->vm->root.bo, f, true); -- cgit v1.2.3 From 44e121fbf155edc93eceaae273038afbf90f52a7 Mon Sep 17 00:00:00 2001 From: Philip Yang Date: Sun, 27 Mar 2022 22:22:44 -0400 Subject: drm/amdgpu: Add tlb_cb for unlocked update MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Flush TLB needs wait for GPU update fence done. MMU notify callback to unmap range from GPUs uses unlocked GPU page table update, so add tlb_cb to unlocked update fence to increase vm->tlb_seq. Signed-off-by: Philip Yang Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 48f326609976..683b08f756ed 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -927,7 +927,7 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, r = vm->update_funcs->commit(¶ms, fence); - if (!unlocked && (!(flags & AMDGPU_PTE_VALID) || params.table_freed)) { + if (!(flags & AMDGPU_PTE_VALID) || params.table_freed) { tlb_cb->vm = vm; if (!fence || !*fence || dma_fence_add_callback(*fence, &tlb_cb->cb, -- cgit v1.2.3 From e84c8ab52dc76b09cb1c8885950ad4eaaf340bf6 Mon Sep 17 00:00:00 2001 From: Haowen Bai Date: Thu, 24 Mar 2022 17:46:02 +0800 Subject: drm/amd/display: Fix pointer dereferenced before checking The value actual_pix_clk_100Hz is dereferencing pointer pix_clk_params before pix_clk_params is being null checked. Fix this by assigning pix_clk_params->requested_pix_clk_100hz to actual_pix_clk_100Hz only if pix_clk_params is not NULL, otherwise just NULL. Signed-off-by: Haowen Bai Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c index cc5128e67daf..c36f8e829344 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c @@ -1254,7 +1254,7 @@ static uint32_t dcn3_get_pix_clk_dividers( struct pixel_clk_params *pix_clk_params, struct pll_settings *pll_settings) { - unsigned long long actual_pix_clk_100Hz = pix_clk_params->requested_pix_clk_100hz; + unsigned long long actual_pix_clk_100Hz = pix_clk_params ? pix_clk_params->requested_pix_clk_100hz : 0; struct dce110_clk_src *clk_src; clk_src = TO_DCE110_CLK_SRC(cs); -- cgit v1.2.3 From 6452c544fbab3030f81229d677c15b9e52d0d934 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 26 Mar 2022 16:37:26 -0700 Subject: drm/amd/display: DCN3.1: don't mark as kernel-doc MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit There is no need for this one static function to be marked as kernel-doc notation. Avoid this doc build warning: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst * Enable CRTC Fixes: 110d3968fe95 ("drm/amd/display: Add DCN3.1 OPTC") Signed-off-by: Randy Dunlap Cc: Alex Deucher Cc: Nicholas Kazlauskas Cc: amd-gfx@lists.freedesktop.org Cc: Christian König Cc: "Pan, Xinhui" Cc: Harry Wentland Cc: Leo Li Cc: Rodrigo Siqueira Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c index e05527a3a8ba..10f897b1cb63 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c @@ -91,8 +91,7 @@ static void optc31_set_odm_combine(struct timing_generator *optc, int *opp_id, i optc1->opp_count = opp_cnt; } -/** - * Enable CRTC +/* * Enable CRTC - call ASIC Control Object to enable Timing generator. */ static bool optc31_enable_crtc(struct timing_generator *optc) -- cgit v1.2.3 From 8f3286942f80650b502532870d7558d206b72484 Mon Sep 17 00:00:00 2001 From: Shirish S Date: Fri, 11 Mar 2022 20:30:17 +0530 Subject: amd/display: set backlight only if required [Why] comparing pwm bl values (coverted) with user brightness(converted) levels in commit_tail leads to continuous setting of backlight via dmub as they don't to match. This leads overdrive in queuing of commands to DMCU that sometimes lead to depending on load on DMCU fw: "[drm:dc_dmub_srv_wait_idle] *ERROR* Error waiting for DMUB idle: status=3" [How] Store last successfully set backlight value and compare with it instead of pwm reads which is not what we should compare with. Signed-off-by: Shirish S Reviewed-by: Harry Wentland Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 7 ++++--- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h | 6 ++++++ 2 files changed, 10 insertions(+), 3 deletions(-) (limited to 'drivers/gpu') 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 35e714d28fee..a6c3e1d74124 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -3973,7 +3973,7 @@ static u32 convert_brightness_to_user(const struct amdgpu_dm_backlight_caps *cap max - min); } -static int amdgpu_dm_backlight_set_level(struct amdgpu_display_manager *dm, +static void amdgpu_dm_backlight_set_level(struct amdgpu_display_manager *dm, int bl_idx, u32 user_brightness) { @@ -4004,7 +4004,8 @@ static int amdgpu_dm_backlight_set_level(struct amdgpu_display_manager *dm, DRM_DEBUG("DM: Failed to update backlight on eDP[%d]\n", bl_idx); } - return rc ? 0 : 1; + if (rc) + dm->actual_brightness[bl_idx] = user_brightness; } static int amdgpu_dm_backlight_update_status(struct backlight_device *bd) @@ -9948,7 +9949,7 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) /* restore the backlight level */ for (i = 0; i < dm->num_of_edps; i++) { if (dm->backlight_dev[i] && - (amdgpu_dm_backlight_get_level(dm, i) != dm->brightness[i])) + (dm->actual_brightness[i] != dm->brightness[i])) amdgpu_dm_backlight_set_level(dm, i, dm->brightness[i]); } #endif 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 6a908d736d6a..7e44b0429448 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -540,6 +540,12 @@ struct amdgpu_display_manager { * cached backlight values. */ u32 brightness[AMDGPU_DM_MAX_NUM_EDP]; + /** + * @actual_brightness: + * + * last successfully applied backlight values. + */ + u32 actual_brightness[AMDGPU_DM_MAX_NUM_EDP]; }; enum dsc_clock_force_state { -- cgit v1.2.3 From 55a2d21bba3db088f25d124c76d8bbaa2f4f5771 Mon Sep 17 00:00:00 2001 From: Christian König Date: Fri, 25 Mar 2022 10:22:36 +0100 Subject: drm/amdgpu: fix some kerneldoc in the VM code v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix two incorrect kerneldocs for the recent VM code changes. v2: fix one more typo Signed-off-by: Christian König Reported-by: kernel test robot Reported-by: Stephen Rothwell Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 683b08f756ed..9992a7311387 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -89,7 +89,7 @@ struct amdgpu_prt_cb { }; /** - * amdgpu_vm_tlb_seq_cb - Helper to increment the TLB flush sequence + * struct amdgpu_vm_tlb_seq_cb - Helper to increment the TLB flush sequence */ struct amdgpu_vm_tlb_seq_cb { /** diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c index a821ada5f8ca..958d7ed97882 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_pt.c @@ -620,7 +620,7 @@ error_free_pt: } /** - * amdgpu_vm_free_table - fre one PD/PT + * amdgpu_vm_pt_free - free one PD/PT * * @entry: PDE to free */ -- cgit v1.2.3 From 4499c90e902f399d69248cb887b222fb3d342d20 Mon Sep 17 00:00:00 2001 From: Christian König Date: Fri, 11 Mar 2022 14:10:13 +0100 Subject: drm/amdgpu: fix incorrect size printing in error msg MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit That are bytes not pages. Signed-off-by: Christian König Acked-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 940752488330..a00022b6ee5b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -472,7 +472,7 @@ static bool amdgpu_bo_validate_size(struct amdgpu_device *adev, fail: DRM_DEBUG("BO size %lu > total memory in domain: %llu\n", size, - man->size << PAGE_SHIFT); + man->size); return false; } -- cgit v1.2.3 From 8fde0248a32d29228520b876b8b27d0c44133734 Mon Sep 17 00:00:00 2001 From: Philip Yang Date: Fri, 25 Mar 2022 20:17:28 -0400 Subject: drm/amdkfd: Use atomic64_t type for pdd->tlb_seq MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit To support multi-thread update page table. Signed-off-by: Philip Yang Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_priv.h | 2 +- drivers/gpu/drm/amd/amdkfd/kfd_process.c | 10 +++++++--- 2 files changed, 8 insertions(+), 4 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index 945982a5d688..e1b7e6afa920 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -705,7 +705,7 @@ struct kfd_process_device { /* VM context for GPUVM allocations */ struct file *drm_file; void *drm_priv; - uint64_t tlb_seq; + atomic64_t tlb_seq; /* GPUVM allocations storage */ struct idr alloc_idr; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index ac8123c1ee8f..9e82d7aa67fa 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -1560,7 +1560,7 @@ int kfd_process_device_init_vm(struct kfd_process_device *pdd, return ret; } pdd->drm_priv = drm_file->private_data; - pdd->tlb_seq = 0; + atomic64_set(&pdd->tlb_seq, 0); ret = kfd_process_device_reserve_ib_mem(pdd); if (ret) @@ -1954,10 +1954,14 @@ void kfd_flush_tlb(struct kfd_process_device *pdd, enum TLB_FLUSH_TYPE type) uint64_t tlb_seq = amdgpu_vm_tlb_seq(vm); struct kfd_dev *dev = pdd->dev; - if (pdd->tlb_seq == tlb_seq) + /* + * It can be that we race and lose here, but that is extremely unlikely + * and the worst thing which could happen is that we flush the changes + * into the TLB once more which is harmless. + */ + if (atomic64_xchg(&pdd->tlb_seq, tlb_seq) == tlb_seq) return; - pdd->tlb_seq = tlb_seq; if (dev->dqm->sched_policy == KFD_SCHED_POLICY_NO_HWS) { /* Nothing to flush until a VMID is assigned, which * only happens when the first queue is created. -- cgit v1.2.3 From 60044748f2c816aad2b5f95c46774ce0aaccf4ca Mon Sep 17 00:00:00 2001 From: Kent Russell Date: Fri, 25 Mar 2022 10:11:36 -0400 Subject: drm/amdgpu: Use switch case for unique_id To ease readability, use switch to set unique_id as supported for the supported IP_VERSIONs, and set it to unsupported by default for all other ASICs. This makes it easier to add IP_VERSIONs later on, and makes it obvious that it is not supported by default, instead of the current logic that assumes that it is supported unless it is not one of the specified IP_VERSIONs. v2: Rebase onto previous IP_VERSION change Signed-off-by: Kent Russell Reviewed-by: Alex Deucher Reviewed-by: Kevin Wang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/amdgpu_pm.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c index 9ce597ded31d..4151db2678fb 100644 --- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c @@ -1988,11 +1988,16 @@ static int default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_ if (adev->flags & AMD_IS_APU) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(unique_id)) { - if (gc_ver != IP_VERSION(9, 0, 1) && - gc_ver != IP_VERSION(9, 4, 0) && - gc_ver != IP_VERSION(9, 4, 1) && - gc_ver != IP_VERSION(9, 4, 2)) + switch (gc_ver) { + case IP_VERSION(9, 0, 1): + case IP_VERSION(9, 4, 0): + case IP_VERSION(9, 4, 1): + case IP_VERSION(9, 4, 2): + *states = ATTR_STATE_SUPPORTED; + break; + default: *states = ATTR_STATE_UNSUPPORTED; + } } else if (DEVICE_ATTR_IS(pp_features)) { if (adev->flags & AMD_IS_APU || gc_ver < IP_VERSION(9, 0, 0)) *states = ATTR_STATE_UNSUPPORTED; -- cgit v1.2.3 From 6e5d64e3eaf663f7564297412575074741421444 Mon Sep 17 00:00:00 2001 From: Kent Russell Date: Tue, 29 Mar 2022 09:42:58 -0400 Subject: drm/amdgpu: Add UNIQUE_ID to MetricsMember_t This will allow us to use the generic *_get_metrics_data functions for ASICs that support unique_id Signed-off-by: Kent Russell Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h index ef57b6089c69..46e34ed8a3c8 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h @@ -1333,6 +1333,8 @@ typedef enum { METRICS_VOLTAGE_VDDGFX, METRICS_SS_APU_SHARE, METRICS_SS_DGPU_SHARE, + METRICS_UNIQUE_ID_UPPER32, + METRICS_UNIQUE_ID_LOWER32, } MetricsMember_t; enum smu_cmn2asic_mapping_type { -- cgit v1.2.3 From 4a93d938a4cbebaafa20cb3ca449eb501f118b40 Mon Sep 17 00:00:00 2001 From: Kent Russell Date: Tue, 29 Mar 2022 09:46:11 -0400 Subject: drm/amdgpu: Use metrics data function to get unique_id for Aldebaran This is abstracted well enough in the get_metrics_data function, so use the function Signed-off-by: Kent Russell Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) (limited to 'drivers/gpu') 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 cd81f848d45a..38af648cb857 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c @@ -650,6 +650,12 @@ static int aldebaran_get_smu_metrics_data(struct smu_context *smu, case METRICS_THROTTLER_STATUS: *value = metrics->ThrottlerStatus; break; + case METRICS_UNIQUE_ID_UPPER32: + *value = metrics->PublicSerialNumUpper32; + break; + case METRICS_UNIQUE_ID_LOWER32: + *value = metrics->PublicSerialNumLower32; + break; default: *value = UINT_MAX; break; @@ -1614,16 +1620,12 @@ static void aldebaran_i2c_control_fini(struct smu_context *smu) static void aldebaran_get_unique_id(struct smu_context *smu) { struct amdgpu_device *adev = smu->adev; - SmuMetrics_t *metrics = smu->smu_table.metrics_table; uint32_t upper32 = 0, lower32 = 0; - int ret; - ret = smu_cmn_get_metrics_table(smu, NULL, false); - if (ret) + if (aldebaran_get_smu_metrics_data(smu, METRICS_UNIQUE_ID_UPPER32, &upper32)) + goto out; + if (aldebaran_get_smu_metrics_data(smu, METRICS_UNIQUE_ID_LOWER32, &lower32)) goto out; - - upper32 = metrics->PublicSerialNumUpper32; - lower32 = metrics->PublicSerialNumLower32; out: adev->unique_id = ((uint64_t)upper32 << 32) | lower32; -- cgit v1.2.3 From ebd9c071d29e56b21ef5155757e0da9926d8d814 Mon Sep 17 00:00:00 2001 From: Kent Russell Date: Wed, 9 Feb 2022 19:56:46 -0500 Subject: drm/amdgpu: Add unique_id support for sienna cichlid This is being added to SMU Metrics, so add the required tie-ins in the kernel. Also create the corresponding unique_id sysfs file. v2: Add FW version check, remove SMU mutex v3: Fix style warning v4: Add MP1 IP_VERSION check to FW version check Signed-off-by: Kent Russell Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/amdgpu_pm.c | 1 + .../inc/pmfw_if/smu11_driver_if_sienna_cichlid.h | 13 +++++++-- .../drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c | 33 ++++++++++++++++++++++ 3 files changed, 45 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c index 4151db2678fb..4a9aabc16fbc 100644 --- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c @@ -1993,6 +1993,7 @@ static int default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_ case IP_VERSION(9, 4, 0): case IP_VERSION(9, 4, 1): case IP_VERSION(9, 4, 2): + case IP_VERSION(10, 3, 0): *states = ATTR_STATE_SUPPORTED; break; default: diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu11_driver_if_sienna_cichlid.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu11_driver_if_sienna_cichlid.h index 3e4a314ef925..5831145646e6 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu11_driver_if_sienna_cichlid.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu11_driver_if_sienna_cichlid.h @@ -1419,8 +1419,11 @@ typedef struct { uint8_t PcieRate ; uint8_t PcieWidth ; uint16_t AverageGfxclkFrequencyTarget; - uint16_t Padding16_2; + uint32_t PublicSerialNumLower32; + uint32_t PublicSerialNumUpper32; + + uint16_t Padding16_2; } SmuMetrics_t; typedef struct { @@ -1476,8 +1479,11 @@ typedef struct { uint8_t PcieRate ; uint8_t PcieWidth ; uint16_t AverageGfxclkFrequencyTarget; - uint16_t Padding16_2; + uint32_t PublicSerialNumLower32; + uint32_t PublicSerialNumUpper32; + + uint16_t Padding16_2; } SmuMetrics_V2_t; typedef struct { @@ -1535,6 +1541,9 @@ typedef struct { uint8_t PcieWidth; uint16_t AverageGfxclkFrequencyTarget; + uint32_t PublicSerialNumLower32; + uint32_t PublicSerialNumUpper32; + } SmuMetrics_V3_t; typedef struct { diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c index 38f04836c82f..b2f3d80e5945 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c @@ -715,6 +715,16 @@ static int sienna_cichlid_get_smu_metrics_data(struct smu_context *smu, *value = use_metrics_v3 ? metrics_v3->CurrFanSpeed : use_metrics_v2 ? metrics_v2->CurrFanSpeed : metrics->CurrFanSpeed; break; + case METRICS_UNIQUE_ID_UPPER32: + *value = use_metrics_v3 ? metrics_v3->PublicSerialNumUpper32 : + use_metrics_v2 ? metrics_v2->PublicSerialNumUpper32 : + metrics->PublicSerialNumUpper32; + break; + case METRICS_UNIQUE_ID_LOWER32: + *value = use_metrics_v3 ? metrics_v3->PublicSerialNumLower32 : + use_metrics_v2 ? metrics_v2->PublicSerialNumLower32 : + metrics->PublicSerialNumLower32; + break; default: *value = UINT_MAX; break; @@ -1773,6 +1783,28 @@ static int sienna_cichlid_read_sensor(struct smu_context *smu, return ret; } +static void sienna_cichlid_get_unique_id(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + uint32_t upper32 = 0, lower32 = 0; + + /* Only supported as of version 0.58.83.0 and only on Sienna Cichlid */ + if (smu->smc_fw_version < 0x3A5300 || + smu->adev->ip_versions[MP1_HWIP][0] != IP_VERSION(11, 0, 7)) + return; + + if (sienna_cichlid_get_smu_metrics_data(smu, METRICS_UNIQUE_ID_UPPER32, &upper32)) + goto out; + if (sienna_cichlid_get_smu_metrics_data(smu, METRICS_UNIQUE_ID_LOWER32, &lower32)) + goto out; + +out: + + adev->unique_id = ((uint64_t)upper32 << 32) | lower32; + if (adev->serial[0] == '\0') + sprintf(adev->serial, "%016llx", adev->unique_id); +} + static int sienna_cichlid_get_uclk_dpm_states(struct smu_context *smu, uint32_t *clocks_in_khz, uint32_t *num_states) { uint32_t num_discrete_levels = 0; @@ -4182,6 +4214,7 @@ static const struct pptable_funcs sienna_cichlid_ppt_funcs = { .get_ecc_info = sienna_cichlid_get_ecc_info, .get_default_config_table_settings = sienna_cichlid_get_default_config_table_settings, .set_config_table = sienna_cichlid_set_config_table, + .get_unique_id = sienna_cichlid_get_unique_id, }; void sienna_cichlid_set_ppt_funcs(struct smu_context *smu) -- cgit v1.2.3 From 851c5f242a03d7e9b00d2d1537d959e54fa101e8 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Sat, 26 Mar 2022 19:24:47 -0100 Subject: drm/amd/display: detach fpu operations from dcn10_validate_bandwidth in calcs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit dcn10_validate_bandwidth is only used on dcn10 files, but is declared in dcn_calcs files. Rename dcn10_* to dcn_* in calcs, remove DC_FP_* wrapper inside DML folder and create an specific dcn10_validate_bandwidth in dcn10_resources that calls dcn_validate_bandwidth and properly wraps that FPU function with DC_FP_* macro. Acked-by: Christian König Reviewed-by: Rodrigo Siqueira Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c | 14 ++++++++++++++ drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c | 5 +---- drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h | 2 +- 3 files changed, 16 insertions(+), 5 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c index 4048908dd265..1587a060b55a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c @@ -1141,6 +1141,20 @@ static void dcn10_destroy_resource_pool(struct resource_pool **pool) *pool = NULL; } +static bool dcn10_validate_bandwidth( + struct dc *dc, + struct dc_state *context, + bool fast_validate) +{ + bool voltage_supported; + + DC_FP_START(); + voltage_supported = dcn_validate_bandwidth(dc, context, fast_validate); + DC_FP_END(); + + return voltage_supported; +} + static enum dc_status dcn10_validate_plane(const struct dc_plane_state *plane_state, struct dc_caps *caps) { if (plane_state->format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN diff --git a/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c b/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c index e447c74be713..c25023f7d604 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c +++ b/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c @@ -764,7 +764,7 @@ static unsigned int get_highest_allowed_voltage_level(uint32_t chip_family, return 4; } -bool dcn10_validate_bandwidth( +bool dcn_validate_bandwidth( struct dc *dc, struct dc_state *context, bool fast_validate) @@ -790,7 +790,6 @@ bool dcn10_validate_bandwidth( dcn_bw_sync_calcs_and_dml(dc); memset(v, 0, sizeof(*v)); - DC_FP_START(); v->sr_exit_time = dc->dcn_soc->sr_exit_time; v->sr_enter_plus_exit_time = dc->dcn_soc->sr_enter_plus_exit_time; @@ -1323,8 +1322,6 @@ bool dcn10_validate_bandwidth( bw_limit = dc->dcn_soc->percent_disp_bw_limit * v->fabric_and_dram_bandwidth_vmax0p9; bw_limit_pass = (v->total_data_read_bandwidth / 1000.0) < bw_limit; - DC_FP_END(); - PERFORMANCE_TRACE_END(); BW_VAL_TRACE_FINISH(); diff --git a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h index 337c0161e72d..806f3041db14 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h @@ -619,7 +619,7 @@ struct dcn_ip_params { }; extern const struct dcn_ip_params dcn10_ip_defaults; -bool dcn10_validate_bandwidth( +bool dcn_validate_bandwidth( struct dc *dc, struct dc_state *context, bool fast_validate); -- cgit v1.2.3 From 9696679bf7ac40a8fb6a488a75bd66d4414cd3c3 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Sat, 26 Mar 2022 19:24:48 -0100 Subject: drm/amd/display: remove DC_FP_* wrapper from dml folder MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit FPU documentation states that developers must not use DC_FP_START/END inside dml files, but use this macro to wrap calls to FPU functions in dc folder (outside dml folder). Therefore, this patch removes DC_FP_* wrappers from dml folder and wraps calls for these FPU operations outside dml, as required. Acked-by: Christian König Reviewed-by: Rodrigo Siqueira Signed-off-by: Melissa Wen Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 10 ++++++++-- drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c | 2 ++ drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c | 14 -------------- drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c | 2 -- 4 files changed, 10 insertions(+), 18 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index e4247740ac12..50820e79d3c4 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c @@ -2983,8 +2983,11 @@ void dcn10_prepare_bandwidth( true); dcn10_stereo_hw_frame_pack_wa(dc, context); - if (dc->debug.pplib_wm_report_mode == WM_REPORT_OVERRIDE) + if (dc->debug.pplib_wm_report_mode == WM_REPORT_OVERRIDE) { + DC_FP_START(); dcn_bw_notify_pplib_of_wm_ranges(dc); + DC_FP_END(); + } if (dc->debug.sanity_checks) hws->funcs.verify_allow_pstate_change_high(dc); @@ -3017,8 +3020,11 @@ void dcn10_optimize_bandwidth( dcn10_stereo_hw_frame_pack_wa(dc, context); - if (dc->debug.pplib_wm_report_mode == WM_REPORT_OVERRIDE) + if (dc->debug.pplib_wm_report_mode == WM_REPORT_OVERRIDE) { + DC_FP_START(); dcn_bw_notify_pplib_of_wm_ranges(dc); + DC_FP_END(); + } if (dc->debug.sanity_checks) hws->funcs.verify_allow_pstate_change_high(dc); diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c index 1587a060b55a..bca049b2f867 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c @@ -1506,6 +1506,7 @@ static bool dcn10_resource_construct( && pool->base.pp_smu->rv_funcs.set_pme_wa_enable != NULL) dc->debug.az_endpoint_mute_only = false; + DC_FP_START(); if (!dc->debug.disable_pplib_clock_request) dcn_bw_update_from_pplib(dc); dcn_bw_sync_calcs_and_dml(dc); @@ -1513,6 +1514,7 @@ static bool dcn10_resource_construct( dc->res_pool = &pool->base; dcn_bw_notify_pplib_of_wm_ranges(dc); } + DC_FP_END(); { struct irq_service_init_data init_data; diff --git a/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c b/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c index c25023f7d604..db3b16b77034 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c +++ b/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c @@ -639,7 +639,6 @@ static bool dcn_bw_apply_registry_override(struct dc *dc) { bool updated = false; - DC_FP_START(); if ((int)(dc->dcn_soc->sr_exit_time * 1000) != dc->debug.sr_exit_time_ns && dc->debug.sr_exit_time_ns) { updated = true; @@ -675,7 +674,6 @@ static bool dcn_bw_apply_registry_override(struct dc *dc) dc->dcn_soc->dram_clock_change_latency = dc->debug.dram_clock_change_latency_ns / 1000.0; } - DC_FP_END(); return updated; } @@ -1492,8 +1490,6 @@ void dcn_bw_update_from_pplib(struct dc *dc) res = dm_pp_get_clock_levels_by_type_with_voltage( ctx, DM_PP_CLOCK_TYPE_FCLK, &fclks); - DC_FP_START(); - if (res) res = verify_clock_values(&fclks); @@ -1523,13 +1519,9 @@ void dcn_bw_update_from_pplib(struct dc *dc) } else BREAK_TO_DEBUGGER(); - DC_FP_END(); - res = dm_pp_get_clock_levels_by_type_with_voltage( ctx, DM_PP_CLOCK_TYPE_DCFCLK, &dcfclks); - DC_FP_START(); - if (res) res = verify_clock_values(&dcfclks); @@ -1540,8 +1532,6 @@ void dcn_bw_update_from_pplib(struct dc *dc) dc->dcn_soc->dcfclkv_max0p9 = dcfclks.data[dcfclks.num_levels - 1].clocks_in_khz / 1000.0; } else BREAK_TO_DEBUGGER(); - - DC_FP_END(); } void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc) @@ -1556,11 +1546,9 @@ void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc) if (!pp || !pp->set_wm_ranges) return; - DC_FP_START(); min_fclk_khz = dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65 * 1000000 / 32; min_dcfclk_khz = dc->dcn_soc->dcfclkv_min0p65 * 1000; socclk_khz = dc->dcn_soc->socclk * 1000; - DC_FP_END(); /* Now notify PPLib/SMU about which Watermarks sets they should select * depending on DPM state they are in. And update BW MGR GFX Engine and @@ -1611,7 +1599,6 @@ void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc) void dcn_bw_sync_calcs_and_dml(struct dc *dc) { - DC_FP_START(); DC_LOG_BANDWIDTH_CALCS("sr_exit_time: %f ns\n" "sr_enter_plus_exit_time: %f ns\n" "urgent_latency: %f ns\n" @@ -1800,5 +1787,4 @@ void dcn_bw_sync_calcs_and_dml(struct dc *dc) dc->dml.ip.bug_forcing_LC_req_same_size_fixed = dc->dcn_ip->bug_forcing_luma_and_chroma_request_to_same_size_fixed == dcn_bw_yes; dc->dml.ip.dcfclk_cstate_latency = dc->dcn_ip->dcfclk_cstate_latency; - DC_FP_END(); } diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c index f93af45aeab4..f79dd40f8d81 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c @@ -1302,9 +1302,7 @@ int dcn20_populate_dml_pipes_from_context( } /* populate writeback information */ - DC_FP_START(); dc->res_pool->funcs->populate_dml_writeback_from_context(dc, res_ctx, pipes); - DC_FP_END(); return pipe_cnt; } -- cgit v1.2.3 From 5e9c445172fde2ee761af91ad302e41041de6be8 Mon Sep 17 00:00:00 2001 From: Kent Russell Date: Wed, 30 Mar 2022 09:27:41 -0400 Subject: drm/amdgpu: Fix unique_id references for Sienna Cichlid Since unique_id is only supported in PMFW 0x3A5300 and higher, we will only be able to use it inside Smu_Metrics_V3_t, which requires PMFW 0x3A4900 and higher. Remove the unique_id/serial_number references from the v1 and v2 tables to avoid any confusion, and return 0 if metrics_v1 or metrics_v2 are used to try to get the unique_id/serial_number. Signed-off-by: Kent Russell Reviewed-by: Lijo Lazar Signed-off-by: Alex Deucher --- .../amd/pm/swsmu/inc/pmfw_if/smu11_driver_if_sienna_cichlid.h | 6 ------ drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c | 10 ++++------ 2 files changed, 4 insertions(+), 12 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu11_driver_if_sienna_cichlid.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu11_driver_if_sienna_cichlid.h index 5831145646e6..08f0bb2af5d2 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu11_driver_if_sienna_cichlid.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu11_driver_if_sienna_cichlid.h @@ -1420,9 +1420,6 @@ typedef struct { uint8_t PcieWidth ; uint16_t AverageGfxclkFrequencyTarget; - uint32_t PublicSerialNumLower32; - uint32_t PublicSerialNumUpper32; - uint16_t Padding16_2; } SmuMetrics_t; @@ -1480,9 +1477,6 @@ typedef struct { uint8_t PcieWidth ; uint16_t AverageGfxclkFrequencyTarget; - uint32_t PublicSerialNumLower32; - uint32_t PublicSerialNumUpper32; - uint16_t Padding16_2; } SmuMetrics_V2_t; diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c index b2f3d80e5945..ab3e9d8b831e 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c @@ -716,14 +716,12 @@ static int sienna_cichlid_get_smu_metrics_data(struct smu_context *smu, use_metrics_v2 ? metrics_v2->CurrFanSpeed : metrics->CurrFanSpeed; break; case METRICS_UNIQUE_ID_UPPER32: - *value = use_metrics_v3 ? metrics_v3->PublicSerialNumUpper32 : - use_metrics_v2 ? metrics_v2->PublicSerialNumUpper32 : - metrics->PublicSerialNumUpper32; + /* Only supported in 0x3A5300+, metrics_v3 requires 0x3A4900+ */ + *value = use_metrics_v3 ? metrics_v3->PublicSerialNumUpper32 : 0; break; case METRICS_UNIQUE_ID_LOWER32: - *value = use_metrics_v3 ? metrics_v3->PublicSerialNumLower32 : - use_metrics_v2 ? metrics_v2->PublicSerialNumLower32 : - metrics->PublicSerialNumLower32; + /* Only supported in 0x3A5300+, metrics_v3 requires 0x3A4900+ */ + *value = use_metrics_v3 ? metrics_v3->PublicSerialNumLower32 : 0; break; default: *value = UINT_MAX; -- cgit v1.2.3 From 058497e1f596985dd21910eae7b4a2b39f94ef23 Mon Sep 17 00:00:00 2001 From: Ruili Ji Date: Thu, 31 Mar 2022 12:19:00 +0800 Subject: drm/amdgpu: fix incorrect GCR_GENERAL_CNTL address gfx10.3.3/gfx10.3.6/gfx10.3.7 shall use 0x1580 address for GCR_GENERAL_CNTL Acked-by: Prike Liang Acked-by: Yifan Zhang Reviewed-by: Aaron Liu Signed-off-by: Ruili Ji Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index a98b78e0b507..9426e252d8aa 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -3293,7 +3293,7 @@ static const struct soc15_reg_golden golden_settings_gc_10_3_3[] = SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG3, 0xffffffff, 0x00000280), SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG4, 0xffffffff, 0x00800000), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG, 0x0c1807ff, 0x00000242), - SOC15_REG_GOLDEN_VALUE(GC, 0, mmGCR_GENERAL_CNTL, 0x1ff1ffff, 0x00000500), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGCR_GENERAL_CNTL_Vangogh, 0x1ff1ffff, 0x00000500), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL1_PIPE_STEER, 0x000000ff, 0x000000e4), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2_PIPE_STEER_0, 0x77777777, 0x32103210), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2_PIPE_STEER_1, 0x77777777, 0x32103210), @@ -3429,7 +3429,7 @@ static const struct soc15_reg_golden golden_settings_gc_10_3_6[] = SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG3, 0xffffffff, 0x00000280), SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG4, 0xffffffff, 0x00800000), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG, 0x0c1807ff, 0x00000042), - SOC15_REG_GOLDEN_VALUE(GC, 0, mmGCR_GENERAL_CNTL, 0x1ff1ffff, 0x00000500), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGCR_GENERAL_CNTL_Vangogh, 0x1ff1ffff, 0x00000500), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL1_PIPE_STEER, 0x000000ff, 0x00000044), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2_PIPE_STEER_0, 0x77777777, 0x32103210), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2_PIPE_STEER_1, 0x77777777, 0x32103210), @@ -3454,7 +3454,7 @@ static const struct soc15_reg_golden golden_settings_gc_10_3_7[] = { SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG3, 0xffffffff, 0x00000280), SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG4, 0xffffffff, 0x00800000), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG, 0x0c1807ff, 0x00000041), - SOC15_REG_GOLDEN_VALUE(GC, 0, mmGCR_GENERAL_CNTL, 0x1ff1ffff, 0x00000500), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGCR_GENERAL_CNTL_Vangogh, 0x1ff1ffff, 0x00000500), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL1_PIPE_STEER, 0x000000ff, 0x000000e4), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2_PIPE_STEER_0, 0x77777777, 0x32103210), SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2_PIPE_STEER_1, 0x77777777, 0x32103210), -- cgit v1.2.3 From cf8cc382aa7c60f25386a8515b6a5e74aef9d0aa Mon Sep 17 00:00:00 2001 From: Ma Jun Date: Wed, 23 Mar 2022 21:16:19 +0800 Subject: drm/amdgpu: Sync up header and implementation to use the same parameter names MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Sync up header and implementation to use the same parameter names in function amdgpu_ring_init. ring_size -> max_dw, prio -> hw_prio Reviewed-by: Guchun Chen Reviewed-by: Christian König Signed-off-by: Ma Jun Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h index 5320bb0883d8..317d80209e95 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h @@ -300,8 +300,8 @@ void amdgpu_ring_generic_pad_ib(struct amdgpu_ring *ring, struct amdgpu_ib *ib); void amdgpu_ring_commit(struct amdgpu_ring *ring); void amdgpu_ring_undo(struct amdgpu_ring *ring); int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring, - unsigned int ring_size, struct amdgpu_irq_src *irq_src, - unsigned int irq_type, unsigned int prio, + unsigned int max_dw, struct amdgpu_irq_src *irq_src, + unsigned int irq_type, unsigned int hw_prio, atomic_t *sched_score); void amdgpu_ring_fini(struct amdgpu_ring *ring); void amdgpu_ring_emit_reg_write_reg_wait_helper(struct amdgpu_ring *ring, -- cgit v1.2.3 From e45422695c196dbc665a95526c85ff4b8752aff2 Mon Sep 17 00:00:00 2001 From: Lee Jones Date: Thu, 31 Mar 2022 13:21:17 +0100 Subject: drm/amdkfd: Create file descriptor after client is added to smi_clients list MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This ensures userspace cannot prematurely clean-up the client before it is fully initialised which has been proven to cause issues in the past. Cc: Felix Kuehling Cc: Alex Deucher Cc: "Christian König" Cc: "Pan, Xinhui" Cc: David Airlie Cc: Daniel Vetter Cc: amd-gfx@lists.freedesktop.org Cc: dri-devel@lists.freedesktop.org Signed-off-by: Lee Jones Reviewed-by: Felix Kuehling Signed-off-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c | 24 +++++++++++++++--------- 1 file changed, 15 insertions(+), 9 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c index e4beebb1c80a..f2e1d506ba21 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c @@ -247,15 +247,6 @@ int kfd_smi_event_open(struct kfd_dev *dev, uint32_t *fd) return ret; } - ret = anon_inode_getfd(kfd_smi_name, &kfd_smi_ev_fops, (void *)client, - O_RDWR); - if (ret < 0) { - kfifo_free(&client->fifo); - kfree(client); - return ret; - } - *fd = ret; - init_waitqueue_head(&client->wait_queue); spin_lock_init(&client->lock); client->events = 0; @@ -265,5 +256,20 @@ int kfd_smi_event_open(struct kfd_dev *dev, uint32_t *fd) list_add_rcu(&client->list, &dev->smi_clients); spin_unlock(&dev->smi_lock); + ret = anon_inode_getfd(kfd_smi_name, &kfd_smi_ev_fops, (void *)client, + O_RDWR); + if (ret < 0) { + spin_lock(&dev->smi_lock); + list_del_rcu(&client->list); + spin_unlock(&dev->smi_lock); + + synchronize_rcu(); + + kfifo_free(&client->fifo); + kfree(client); + return ret; + } + *fd = ret; + return 0; } -- cgit v1.2.3 From 50e6cb3fd2cde554db646282ea10df7236e6493c Mon Sep 17 00:00:00 2001 From: CHANDAN VURDIGERE NATARAJ Date: Tue, 29 Mar 2022 13:10:31 +0530 Subject: drm/amd/display: Fix by adding FPU protection for dcn30_internal_validate_bw [Why] Below general protection fault observed when WebGL Aquarium is run for longer duration. If drm debug logs are enabled and set to 0x1f then the issue is observed within 10 minutes of run. [ 100.717056] general protection fault, probably for non-canonical address 0x2d33302d32323032: 0000 [#1] PREEMPT SMP NOPTI [ 100.727921] CPU: 3 PID: 1906 Comm: DrmThread Tainted: G W 5.15.30 #12 d726c6a2d6ebe5cf9223931cbca6892f916fe18b [ 100.754419] RIP: 0010:CalculateSwathWidth+0x1f7/0x44f [ 100.767109] Code: 00 00 00 f2 42 0f 11 04 f0 48 8b 85 88 00 00 00 f2 42 0f 10 04 f0 48 8b 85 98 00 00 00 f2 42 0f 11 04 f0 48 8b 45 10 0f 57 c0 42 0f 2a 04 b0 0f 57 c9 f3 43 0f 2a 0c b4 e8 8c e2 f3 ff 48 8b [ 100.781269] RSP: 0018:ffffa9230079eeb0 EFLAGS: 00010246 [ 100.812528] RAX: 2d33302d32323032 RBX: 0000000000000500 RCX: 0000000000000000 [ 100.819656] RDX: 0000000000000001 RSI: ffff99deb712c49c RDI: 0000000000000000 [ 100.826781] RBP: ffffa9230079ef50 R08: ffff99deb712460c R09: ffff99deb712462c [ 100.833907] R10: ffff99deb7124940 R11: ffff99deb7124d70 R12: ffff99deb712ae44 [ 100.841033] R13: 0000000000000001 R14: 0000000000000000 R15: ffffa9230079f0a0 [ 100.848159] FS: 00007af121212640(0000) GS:ffff99deba780000(0000) knlGS:0000000000000000 [ 100.856240] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 100.861980] CR2: 0000209000fe1000 CR3: 000000011b18c000 CR4: 0000000000350ee0 [ 100.869106] Call Trace: [ 100.871555] [ 100.873655] ? asm_sysvec_reschedule_ipi+0x12/0x20 [ 100.878449] CalculateSwathAndDETConfiguration+0x1a3/0x6dd [ 100.883937] dml31_ModeSupportAndSystemConfigurationFull+0x2ce4/0x76da [ 100.890467] ? kallsyms_lookup_buildid+0xc8/0x163 [ 100.895173] ? kallsyms_lookup_buildid+0xc8/0x163 [ 100.899874] ? __sprint_symbol+0x80/0x135 [ 100.903883] ? dm_update_plane_state+0x3f9/0x4d2 [ 100.908500] ? symbol_string+0xb7/0xde [ 100.912250] ? number+0x145/0x29b [ 100.915566] ? vsnprintf+0x341/0x5ff [ 100.919141] ? desc_read_finalized_seq+0x39/0x87 [ 100.923755] ? update_load_avg+0x1b9/0x607 [ 100.927849] ? compute_mst_dsc_configs_for_state+0x7d/0xd5b [ 100.933416] ? fetch_pipe_params+0xa4d/0xd0c [ 100.937686] ? dc_fpu_end+0x3d/0xa8 [ 100.941175] dml_get_voltage_level+0x16b/0x180 [ 100.945619] dcn30_internal_validate_bw+0x10e/0x89b [ 100.950495] ? dcn31_validate_bandwidth+0x68/0x1fc [ 100.955285] ? resource_build_scaling_params+0x98b/0xb8c [ 100.960595] ? dcn31_validate_bandwidth+0x68/0x1fc [ 100.965384] dcn31_validate_bandwidth+0x9a/0x1fc [ 100.970001] dc_validate_global_state+0x238/0x295 [ 100.974703] amdgpu_dm_atomic_check+0x9c1/0xbce [ 100.979235] ? _printk+0x59/0x73 [ 100.982467] drm_atomic_check_only+0x403/0x78b [ 100.986912] drm_mode_atomic_ioctl+0x49b/0x546 [ 100.991358] ? drm_ioctl+0x1c1/0x3b3 [ 100.994936] ? drm_atomic_set_property+0x92a/0x92a [ 100.999725] drm_ioctl_kernel+0xdc/0x149 [ 101.003648] drm_ioctl+0x27f/0x3b3 [ 101.007051] ? drm_atomic_set_property+0x92a/0x92a [ 101.011842] amdgpu_drm_ioctl+0x49/0x7d [ 101.015679] __se_sys_ioctl+0x7c/0xb8 [ 101.015685] do_syscall_64+0x5f/0xb8 [ 101.015690] ? __irq_exit_rcu+0x34/0x96 [How] It calles populate_dml_pipes which uses doubles to initialize. Adding FPU protection avoids context switch and probable loss of vba context as there is potential contention while drm debug logs are enabled. Signed-off-by: CHANDAN VURDIGERE NATARAJ Reviewed-by: Rodrigo Siqueira Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c index 826970f2bd0a..f27262417abe 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c @@ -1750,7 +1750,9 @@ bool dcn31_validate_bandwidth(struct dc *dc, BW_VAL_TRACE_COUNT(); + DC_FP_START(); out = dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, fast_validate); + DC_FP_END(); // Disable fast_validate to set min dcfclk in alculate_wm_and_dlg if (pipe_cnt == 0) -- cgit v1.2.3 From 30671b44aa570a2953aead09999d13e3b5a24d30 Mon Sep 17 00:00:00 2001 From: Christian König Date: Wed, 30 Mar 2022 10:53:15 +0200 Subject: drm/amdgpu: fix TLB flushing during eviction MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Testing the valid bit is not enough to figure out if we need to invalidate the TLB or not. During eviction it is quite likely that we move a BO from VRAM to GTT and update the page tables immediately to the new GTT address. Rework the whole function to get all the necessary parameters directly as value. Signed-off-by: Christian König Reviewed-by: Philip Yang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 63 ++++++++++++++++++---------------- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 15 ++++---- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 18 +++++----- 3 files changed, 48 insertions(+), 48 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 9992a7311387..f0aec04111a3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -793,18 +793,19 @@ static void amdgpu_vm_tlb_seq_cb(struct dma_fence *fence, } /** - * amdgpu_vm_bo_update_mapping - update a mapping in the vm page table + * amdgpu_vm_update_range - update a range in the vm page table * - * @adev: amdgpu_device pointer of the VM - * @bo_adev: amdgpu_device pointer of the mapped BO - * @vm: requested vm + * @adev: amdgpu_device pointer to use for commands + * @vm: the VM to update the range * @immediate: immediate submission in a page fault * @unlocked: unlocked invalidation during MM callback + * @flush_tlb: trigger tlb invalidation after update completed * @resv: fences we need to sync to * @start: start of mapped range * @last: last mapped entry * @flags: flags for the entries * @offset: offset into nodes and pages_addr + * @vram_base: base for vram mappings * @res: ttm_resource to map * @pages_addr: DMA addresses to use for mapping * @fence: optional resulting fence @@ -812,17 +813,14 @@ static void amdgpu_vm_tlb_seq_cb(struct dma_fence *fence, * Fill in the page table entries between @start and @last. * * Returns: - * 0 for success, -EINVAL for failure. + * 0 for success, negative erro code for failure. */ -int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, - struct amdgpu_device *bo_adev, - struct amdgpu_vm *vm, bool immediate, - bool unlocked, struct dma_resv *resv, - uint64_t start, uint64_t last, - uint64_t flags, uint64_t offset, - struct ttm_resource *res, - dma_addr_t *pages_addr, - struct dma_fence **fence) +int amdgpu_vm_update_range(struct amdgpu_device *adev, struct amdgpu_vm *vm, + bool immediate, bool unlocked, bool flush_tlb, + struct dma_resv *resv, uint64_t start, uint64_t last, + uint64_t flags, uint64_t offset, uint64_t vram_base, + struct ttm_resource *res, dma_addr_t *pages_addr, + struct dma_fence **fence) { struct amdgpu_vm_update_params params; struct amdgpu_vm_tlb_seq_cb *tlb_cb; @@ -910,8 +908,7 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, } } else if (flags & (AMDGPU_PTE_VALID | AMDGPU_PTE_PRT)) { - addr = bo_adev->vm_manager.vram_base_offset + - cursor.start; + addr = vram_base + cursor.start; } else { addr = 0; } @@ -927,7 +924,7 @@ int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, r = vm->update_funcs->commit(¶ms, fence); - if (!(flags & AMDGPU_PTE_VALID) || params.table_freed) { + if (flush_tlb || params.table_freed) { tlb_cb->vm = vm; if (!fence || !*fence || dma_fence_add_callback(*fence, &tlb_cb->cb, @@ -1010,9 +1007,10 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, dma_addr_t *pages_addr = NULL; struct ttm_resource *mem; struct dma_fence **last_update; + bool flush_tlb = clear; struct dma_resv *resv; + uint64_t vram_base; uint64_t flags; - struct amdgpu_device *bo_adev = adev; int r; if (clear || !bo) { @@ -1037,14 +1035,18 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, } if (bo) { + struct amdgpu_device *bo_adev; + flags = amdgpu_ttm_tt_pte_flags(adev, bo->tbo.ttm, mem); if (amdgpu_bo_encrypted(bo)) flags |= AMDGPU_PTE_TMZ; bo_adev = amdgpu_ttm_adev(bo->tbo.bdev); + vram_base = bo_adev->vm_manager.vram_base_offset; } else { flags = 0x0; + vram_base = 0; } if (clear || (bo && bo->tbo.base.resv == @@ -1054,7 +1056,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, last_update = &bo_va->last_pt_update; if (!clear && bo_va->base.moved) { - bo_va->base.moved = false; + flush_tlb = true; list_splice_init(&bo_va->valids, &bo_va->invalids); } else if (bo_va->cleared != clear) { @@ -1077,11 +1079,11 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, trace_amdgpu_vm_bo_update(mapping); - r = amdgpu_vm_bo_update_mapping(adev, bo_adev, vm, false, false, - resv, mapping->start, - mapping->last, update_flags, - mapping->offset, mem, - pages_addr, last_update); + r = amdgpu_vm_update_range(adev, vm, false, false, flush_tlb, + resv, mapping->start, mapping->last, + update_flags, mapping->offset, + vram_base, mem, pages_addr, + last_update); if (r) return r; } @@ -1104,6 +1106,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, list_splice_init(&bo_va->invalids, &bo_va->valids); bo_va->cleared = clear; + bo_va->base.moved = false; if (trace_amdgpu_vm_bo_mapping_enabled()) { list_for_each_entry(mapping, &bo_va->valids, list) @@ -1272,10 +1275,10 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev, mapping->start < AMDGPU_GMC_HOLE_START) init_pte_value = AMDGPU_PTE_DEFAULT_ATC; - r = amdgpu_vm_bo_update_mapping(adev, adev, vm, false, false, - resv, mapping->start, - mapping->last, init_pte_value, - 0, NULL, NULL, &f); + r = amdgpu_vm_update_range(adev, vm, false, false, true, resv, + mapping->start, mapping->last, + init_pte_value, 0, 0, NULL, NULL, + &f); amdgpu_vm_free_mapping(adev, vm, mapping, f); if (r) { dma_fence_put(f); @@ -2519,8 +2522,8 @@ bool amdgpu_vm_handle_fault(struct amdgpu_device *adev, u32 pasid, goto error_unlock; } - r = amdgpu_vm_bo_update_mapping(adev, adev, vm, true, false, NULL, addr, - addr, flags, value, NULL, NULL, NULL); + r = amdgpu_vm_update_range(adev, vm, true, false, false, NULL, addr, + addr, flags, value, 0, NULL, NULL, NULL); if (r) goto error_unlock; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index 6b7682fe84f8..1a814fbffff8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -402,15 +402,12 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev, struct amdgpu_vm *vm); void amdgpu_vm_bo_base_init(struct amdgpu_vm_bo_base *base, struct amdgpu_vm *vm, struct amdgpu_bo *bo); -int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, - struct amdgpu_device *bo_adev, - struct amdgpu_vm *vm, bool immediate, - bool unlocked, struct dma_resv *resv, - uint64_t start, uint64_t last, - uint64_t flags, uint64_t offset, - struct ttm_resource *res, - dma_addr_t *pages_addr, - struct dma_fence **fence); +int amdgpu_vm_update_range(struct amdgpu_device *adev, struct amdgpu_vm *vm, + bool immediate, bool unlocked, bool flush_tlb, + struct dma_resv *resv, uint64_t start, uint64_t last, + uint64_t flags, uint64_t offset, uint64_t vram_base, + struct ttm_resource *res, dma_addr_t *pages_addr, + struct dma_fence **fence); int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va, bool clear); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index 27533f6057e0..907b02045824 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -1188,9 +1188,9 @@ svm_range_unmap_from_gpu(struct amdgpu_device *adev, struct amdgpu_vm *vm, pr_debug("[0x%llx 0x%llx]\n", start, last); - return amdgpu_vm_bo_update_mapping(adev, adev, vm, false, true, NULL, - start, last, init_pte_value, 0, - NULL, NULL, fence); + return amdgpu_vm_update_range(adev, vm, false, true, true, NULL, start, + last, init_pte_value, 0, 0, NULL, NULL, + fence); } static int @@ -1277,12 +1277,12 @@ svm_range_map_to_gpu(struct kfd_process_device *pdd, struct svm_range *prange, (last_domain == SVM_RANGE_VRAM_DOMAIN) ? 1 : 0, pte_flags); - r = amdgpu_vm_bo_update_mapping(adev, bo_adev, vm, false, false, - NULL, last_start, - prange->start + i, pte_flags, - last_start - prange->start, - NULL, dma_addr, - &vm->last_update); + r = amdgpu_vm_update_range(adev, vm, false, false, false, NULL, + last_start, prange->start + i, + pte_flags, + last_start - prange->start, + bo_adev->vm_manager.vram_base_offset, + NULL, dma_addr, &vm->last_update); for (j = last_start - prange->start; j <= i; j++) dma_addr[j] |= last_domain; -- cgit v1.2.3 From a68bec2ce7d6d89136b91160c3428caf683acb91 Mon Sep 17 00:00:00 2001 From: Marko Zekovic Date: Mon, 28 Mar 2022 12:37:30 +0200 Subject: amdgpu/pm: Enable sysfs nodes for vclk and dclk for NAVI12 SMI clock measure API is failing on NAVI12, because sysfs node for pp_dpm_vclk is not existing. Enable sysfs node for pp_dpm_vclk for NAVI12. v2: Also enable sysfs node for pp_dpm_dclk. Reviewed-by: Lijo Lazar Signed-off-by: Marko Zekovic Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/amdgpu_pm.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c index 4a9aabc16fbc..ec4ea141bfda 100644 --- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c @@ -2007,11 +2007,13 @@ static int default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_ *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pp_dpm_vclk)) { if (!(gc_ver == IP_VERSION(10, 3, 1) || - gc_ver == IP_VERSION(10, 3, 0))) + gc_ver == IP_VERSION(10, 3, 0) || + gc_ver == IP_VERSION(10, 1, 2))) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pp_dpm_dclk)) { if (!(gc_ver == IP_VERSION(10, 3, 1) || - gc_ver == IP_VERSION(10, 3, 0))) + gc_ver == IP_VERSION(10, 3, 0) || + gc_ver == IP_VERSION(10, 1, 2))) *states = ATTR_STATE_UNSUPPORTED; } else if (DEVICE_ATTR_IS(pp_power_profile_mode)) { if (amdgpu_dpm_get_power_profile_mode(adev, NULL) == -EOPNOTSUPP) -- cgit v1.2.3 From 7da7b02e97c8e4332a497a179575db6892cbeca5 Mon Sep 17 00:00:00 2001 From: Aashish Sharma Date: Thu, 24 Mar 2022 19:10:31 +0530 Subject: drm/amd/display: Fix unused-but-set-variable warning Fix the kernel test robot warning below: drivers/gpu/drm/amd/amdgpu/../display/dmub/inc/dmub_cmd.h:2893:12: warning: variable 'temp' set but not used [-Wunused-but-set-variable] Replaced the assignment to the unused temp variable with READ_ONCE() macro to flush the writes. READ_ONCE() helps avoid the use of volatile and makes it obvious from the code that the read here is intentional. Also verified on x86 that the generated code is exactly the same as before. Reported-by: kernel test robot Signed-off-by: Aashish Sharma Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'drivers/gpu') 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 9b5db16b2619..95f1c7198dd3 100644 --- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h @@ -2962,9 +2962,7 @@ static inline void dmub_rb_flush_pending(const struct dmub_rb *rb) uint32_t wptr = rb->wrpt; while (rptr != wptr) { - uint64_t volatile *data = (uint64_t volatile *)((uint8_t *)(rb->base_address) + rptr); - //uint64_t volatile *p = (uint64_t volatile *)data; - uint64_t temp; + uint64_t *data = (uint64_t *)((uint8_t *)(rb->base_address) + rptr); uint8_t i; /* Don't remove this. @@ -2972,7 +2970,7 @@ static inline void dmub_rb_flush_pending(const struct dmub_rb *rb) * for this function to be effective. */ for (i = 0; i < DMUB_RB_CMD_SIZE / sizeof(uint64_t); i++) - temp = *data++; + (void)READ_ONCE(*data++); rptr += DMUB_RB_CMD_SIZE; if (rptr >= rb->capacity) -- cgit v1.2.3 From 7e97de3e7f1fb33e8c7057101095d7ddf046077c Mon Sep 17 00:00:00 2001 From: Haowen Bai Date: Fri, 1 Apr 2022 14:28:05 +0800 Subject: drm/amdgpu/vcn: Remove unneeded semicolon report by coccicheck: drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c:1951:2-3: Unneeded semicolon Fixes: c543dcbe4237 ("drm/amdgpu/vcn: Add VCN ras error query support") Signed-off-by: Haowen Bai Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c index 3e1de8c8b372..17d44be58877 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c @@ -1948,7 +1948,7 @@ static uint32_t vcn_v2_6_query_poison_by_instance(struct amdgpu_device *adev, break; default: break; - }; + } if (poison_stat) dev_info(adev->dev, "Poison detected in VCN%d, sub_block%d\n", -- cgit v1.2.3 From 34452ac3038a7dea7e5407c0f06f762412e679f3 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Fri, 25 Mar 2022 11:53:39 -0400 Subject: drm/amdgpu: don't use BACO for reset in S3 Seems to cause a reboots or hangs on some systems. Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1924 Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1953 Fixes: daf8de0874ab5b ("drm/amdgpu: always reset the asic in suspend (v2)") Reviewed-by: Lijo Lazar Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/amdgpu_dpm.c | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/pm/amdgpu_dpm.c b/drivers/gpu/drm/amd/pm/amdgpu_dpm.c index c73fb73e9628..5504d81c77b7 100644 --- a/drivers/gpu/drm/amd/pm/amdgpu_dpm.c +++ b/drivers/gpu/drm/amd/pm/amdgpu_dpm.c @@ -173,6 +173,17 @@ bool amdgpu_dpm_is_baco_supported(struct amdgpu_device *adev) if (!pp_funcs || !pp_funcs->get_asic_baco_capability) return false; + /* Don't use baco for reset in S3. + * This is a workaround for some platforms + * where entering BACO during suspend + * seems to cause reboots or hangs. + * This might be related to the fact that BACO controls + * power to the whole GPU including devices like audio and USB. + * Powering down/up everything may adversely affect these other + * devices. Needs more investigation. + */ + if (adev->in_s3) + return false; mutex_lock(&adev->pm.mutex); -- cgit v1.2.3 From 0f12a22f375400a3fc42b86a0f8c23da530fb0fc Mon Sep 17 00:00:00 2001 From: Philip Yang Date: Fri, 1 Apr 2022 15:30:12 -0400 Subject: drm/amdgpu: Flush TLB after mapping for VG20+XGMI MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit For VG20 + XGMI bridge, all mappings PTEs cache in TC, this may have stall invalid PTEs in TC because one cache line has 8 pages. Need always flush_tlb after updating mapping. Signed-off-by: Philip Yang Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index f0aec04111a3..383242952a20 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -837,6 +837,12 @@ int amdgpu_vm_update_range(struct amdgpu_device *adev, struct amdgpu_vm *vm, goto error_unlock; } + /* Vega20+XGMI where PTEs get inadvertently cached in L2 texture cache, + * heavy-weight flush TLB unconditionally. + */ + flush_tlb |= adev->gmc.xgmi.num_physical_nodes && + adev->ip_versions[GC_HWIP][0] == IP_VERSION(9, 4, 0); + memset(¶ms, 0, sizeof(params)); params.adev = adev; params.vm = vm; -- cgit v1.2.3 From 35b2186b13b66075035fa008e9390d14b331e1b3 Mon Sep 17 00:00:00 2001 From: Eric Bernstein Date: Mon, 21 Mar 2022 10:42:34 -0400 Subject: drm/amd/display: remove assert for odm transition case Remove assert that will hit during odm transition case, since this is a valid case. Tested-by: Daniel Wheeler Reviewed-by: Alvin Lee Acked-by: Tom Chung Signed-off-by: Eric Bernstein Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c | 1 - 1 file changed, 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c index d473708d5399..7802d603f796 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c @@ -1976,7 +1976,6 @@ int dcn20_validate_apply_pipe_split_flags( /*If need split for odm but 4 way split already*/ if (split[i] == 2 && ((pipe->prev_odm_pipe && !pipe->prev_odm_pipe->prev_odm_pipe) || !pipe->next_odm_pipe)) { - ASSERT(0); /* NOT expected yet */ merge[i] = true; /* 4 -> 2 ODM */ } else if (split[i] == 0 && pipe->prev_odm_pipe) { ASSERT(0); /* NOT expected yet */ -- cgit v1.2.3 From 3fe5739db488434bc0368577615ea7275b0f43a5 Mon Sep 17 00:00:00 2001 From: Angus Wang Date: Tue, 22 Mar 2022 15:37:15 -0400 Subject: drm/amd/display: Add flip interval workaround [WHY] Some games experience low FPS issues when FreeSync is on and VSync is toggled to half refresh rate. [HOW] First create a function to determine workaround conditions, which is when we detect 2 or more VSync interrupts between flips and a very short VSync to flip interval. We do the workaround during VSync interrupts and set the v_total_max and min to nominal. We also cleanup after we exit the game. Tested-by: Daniel Wheeler Reviewed-by: Anthony Koo Reviewed-by: Aric Cyr Acked-by: Tom Chung Signed-off-by: Angus Wang Signed-off-by: Alex Deucher --- .../drm/amd/display/modules/freesync/freesync.c | 74 +++++++++++++++++++++- .../gpu/drm/amd/display/modules/inc/mod_freesync.h | 12 ++++ 2 files changed, 85 insertions(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c index bd1d1dc93629..d72566c6928a 100644 --- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c +++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c @@ -29,6 +29,7 @@ #include "dc.h" #include "mod_freesync.h" #include "core_types.h" +#include "dm_services.h" #define MOD_FREESYNC_MAX_CONCURRENT_STREAMS 32 @@ -46,6 +47,10 @@ /* Number of consecutive frames to check before entering/exiting fixed refresh */ #define FIXED_REFRESH_ENTER_FRAME_COUNT 5 #define FIXED_REFRESH_EXIT_FRAME_COUNT 10 +/* Flip interval workaround constants */ +#define VSYNCS_BETWEEN_FLIP_THRESHOLD 2 +#define FREESYNC_CONSEC_FLIP_AFTER_VSYNC 5 +#define FREESYNC_VSYNC_TO_FLIP_DELTA_IN_US 500 struct core_freesync { struct mod_freesync public; @@ -466,6 +471,41 @@ static void apply_fixed_refresh(struct core_freesync *core_freesync, } } +static void determine_flip_interval_workaround_req(struct mod_vrr_params *in_vrr, + unsigned int curr_time_stamp_in_us) +{ + in_vrr->flip_interval.vsync_to_flip_in_us = curr_time_stamp_in_us - + in_vrr->flip_interval.v_update_timestamp_in_us; + + /* Determine conditions for stopping workaround */ + if (in_vrr->flip_interval.flip_interval_workaround_active && + in_vrr->flip_interval.vsyncs_between_flip < VSYNCS_BETWEEN_FLIP_THRESHOLD && + in_vrr->flip_interval.vsync_to_flip_in_us > FREESYNC_VSYNC_TO_FLIP_DELTA_IN_US) { + in_vrr->flip_interval.flip_interval_detect_counter = 0; + in_vrr->flip_interval.program_flip_interval_workaround = true; + in_vrr->flip_interval.flip_interval_workaround_active = false; + } else { + /* Determine conditions for starting workaround */ + if (in_vrr->flip_interval.vsyncs_between_flip >= VSYNCS_BETWEEN_FLIP_THRESHOLD && + in_vrr->flip_interval.vsync_to_flip_in_us < FREESYNC_VSYNC_TO_FLIP_DELTA_IN_US) { + /* Increase flip interval counter we have 2 vsyncs between flips and + * vsync to flip interval is less than 500us + */ + in_vrr->flip_interval.flip_interval_detect_counter++; + if (in_vrr->flip_interval.flip_interval_detect_counter > FREESYNC_CONSEC_FLIP_AFTER_VSYNC) { + /* Start workaround if we detect 5 consecutive instances of the above case */ + in_vrr->flip_interval.program_flip_interval_workaround = true; + in_vrr->flip_interval.flip_interval_workaround_active = true; + } + } else { + /* Reset the flip interval counter if we condition is no longer met */ + in_vrr->flip_interval.flip_interval_detect_counter = 0; + } + } + + in_vrr->flip_interval.vsyncs_between_flip = 0; +} + static bool vrr_settings_require_update(struct core_freesync *core_freesync, struct mod_freesync_config *in_config, unsigned int min_refresh_in_uhz, @@ -1179,6 +1219,9 @@ void mod_freesync_handle_preflip(struct mod_freesync *mod_freesync, in_out_vrr); } + determine_flip_interval_workaround_req(in_out_vrr, + curr_time_stamp_in_us); + } } @@ -1187,6 +1230,7 @@ void mod_freesync_handle_v_update(struct mod_freesync *mod_freesync, struct mod_vrr_params *in_out_vrr) { struct core_freesync *core_freesync = NULL; + unsigned int cur_timestamp_in_us; if ((mod_freesync == NULL) || (stream == NULL) || (in_out_vrr == NULL)) return; @@ -1196,6 +1240,34 @@ void mod_freesync_handle_v_update(struct mod_freesync *mod_freesync, if (in_out_vrr->supported == false) return; + cur_timestamp_in_us = dm_get_timestamp(core_freesync->dc->ctx)/10; + + in_out_vrr->flip_interval.vsyncs_between_flip++; + in_out_vrr->flip_interval.v_update_timestamp_in_us = cur_timestamp_in_us; + + if (in_out_vrr->state == VRR_STATE_ACTIVE_VARIABLE && + (in_out_vrr->flip_interval.flip_interval_workaround_active || + (!in_out_vrr->flip_interval.flip_interval_workaround_active && + in_out_vrr->flip_interval.program_flip_interval_workaround))) { + // set freesync vmin vmax to nominal for workaround + in_out_vrr->adjust.v_total_min = + mod_freesync_calc_v_total_from_refresh( + stream, in_out_vrr->max_refresh_in_uhz); + in_out_vrr->adjust.v_total_max = + in_out_vrr->adjust.v_total_min; + in_out_vrr->flip_interval.program_flip_interval_workaround = false; + in_out_vrr->flip_interval.do_flip_interval_workaround_cleanup = true; + return; + } + + if (in_out_vrr->state != VRR_STATE_ACTIVE_VARIABLE && + in_out_vrr->flip_interval.do_flip_interval_workaround_cleanup) { + in_out_vrr->flip_interval.do_flip_interval_workaround_cleanup = false; + in_out_vrr->flip_interval.flip_interval_detect_counter = 0; + in_out_vrr->flip_interval.vsyncs_between_flip = 0; + in_out_vrr->flip_interval.vsync_to_flip_in_us = 0; + } + /* Below the Range Logic */ /* Only execute if in fullscreen mode */ @@ -1302,7 +1374,7 @@ unsigned long long mod_freesync_calc_field_rate_from_timing( bool mod_freesync_is_valid_range(uint32_t min_refresh_cap_in_uhz, uint32_t max_refresh_cap_in_uhz, - uint32_t nominal_field_rate_in_uhz) + uint32_t nominal_field_rate_in_uhz) { /* Typically nominal refresh calculated can have some fractional part. diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h b/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h index 75a158a2514c..cf6bc9446244 100644 --- a/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h +++ b/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h @@ -105,6 +105,16 @@ struct mod_vrr_params_fixed_refresh { uint32_t frame_counter; }; +struct mod_vrr_params_flip_interval { + bool flip_interval_workaround_active; + bool program_flip_interval_workaround; + bool do_flip_interval_workaround_cleanup; + uint32_t flip_interval_detect_counter; + uint32_t vsyncs_between_flip; + uint32_t vsync_to_flip_in_us; + uint32_t v_update_timestamp_in_us; +}; + struct mod_vrr_params { bool supported; bool send_info_frame; @@ -121,6 +131,8 @@ struct mod_vrr_params { struct mod_vrr_params_fixed_refresh fixed; struct mod_vrr_params_btr btr; + + struct mod_vrr_params_flip_interval flip_interval; }; struct mod_freesync *mod_freesync_create(struct dc *dc); -- cgit v1.2.3 From ec313220c39a195afb1b2948a78f0c88b31cee14 Mon Sep 17 00:00:00 2001 From: Angus Wang Date: Wed, 23 Mar 2022 12:40:16 -0400 Subject: drm/amd/display: Remove underflow IRQ type [WHY] Feature using the underflow IRQ type reverted [HOW] Removed underflow IRQ type entry Tested-by: Daniel Wheeler Reviewed-by: Martin Leung Acked-by: Tom Chung Signed-off-by: Angus Wang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/irq_types.h | 1 - 1 file changed, 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/irq_types.h b/drivers/gpu/drm/amd/display/dc/irq_types.h index 36d8243cfbff..530c2578db40 100644 --- a/drivers/gpu/drm/amd/display/dc/irq_types.h +++ b/drivers/gpu/drm/amd/display/dc/irq_types.h @@ -162,7 +162,6 @@ enum irq_type IRQ_TYPE_VUPDATE = DC_IRQ_SOURCE_VUPDATE1, IRQ_TYPE_VBLANK = DC_IRQ_SOURCE_VBLANK1, IRQ_TYPE_VLINE0 = DC_IRQ_SOURCE_DC1_VLINE0, - IRQ_TYPE_DCUNDERFLOW = DC_IRQ_SOURCE_DC1UNDERFLOW, }; #define DAL_VALID_IRQ_SRC_NUM(src) \ -- cgit v1.2.3 From bb90786fa8d9ab1259c7d38516de5fb5ed145c01 Mon Sep 17 00:00:00 2001 From: Jimmy Kizito Date: Tue, 22 Mar 2022 19:12:47 -0400 Subject: drm/amd/display: Add configuration options for AUX wake work around. [Why] Work around to try to wake unresponsive DP sinks may need to be adjusted for certain sinks. [How] Add options to disable work around or adjust time spent trying to wake unresponsive DPRX. Tested-by: Daniel Wheeler Reviewed-by: Martin Leung Acked-by: Tom Chung Signed-off-by: Jimmy Kizito Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c | 8 +++++++- drivers/gpu/drm/amd/display/dc/dc.h | 17 +++++++++++++++++ 2 files changed, 24 insertions(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c index e4df81dc1dc2..22dabe596dfc 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c @@ -5308,7 +5308,13 @@ static bool retrieve_link_cap(struct dc_link *link) LINK_AUX_DEFAULT_LTTPR_TIMEOUT_PERIOD); /* Try to ensure AUX channel active before proceeding. */ - status = wa_try_to_wake_dprx(link, LINK_AUX_WAKE_TIMEOUT_MS); + if (link->dc->debug.aux_wake_wa.bits.enable_wa) { + uint64_t timeout_ms = link->dc->debug.aux_wake_wa.bits.timeout_ms; + + if (link->dc->debug.aux_wake_wa.bits.use_default_timeout) + timeout_ms = LINK_AUX_WAKE_TIMEOUT_MS; + status = wa_try_to_wake_dprx(link, timeout_ms); + } is_lttpr_present = dp_retrieve_lttpr_cap(link); /* Read DP tunneling information. */ diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index ef286aa30294..e148db770722 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -526,6 +526,22 @@ union dpia_debug_options { uint32_t raw; }; +/* AUX wake work around options + * 0: enable/disable work around + * 1: use default timeout LINK_AUX_WAKE_TIMEOUT_MS + * 15-2: reserved + * 31-16: timeout in ms + */ +union aux_wake_wa_options { + struct { + uint32_t enable_wa : 1; + uint32_t use_default_timeout : 1; + uint32_t rsvd: 14; + uint32_t timeout_ms : 16; + } bits; + uint32_t raw; +}; + struct dc_debug_data { uint32_t ltFailCount; uint32_t i2cErrorCount; @@ -712,6 +728,7 @@ struct dc_debug_options { #endif bool apply_vendor_specific_lttpr_wa; bool extended_blank_optimization; + union aux_wake_wa_options aux_wake_wa; bool ignore_dpref_ss; uint8_t psr_power_use_phy_fsm; }; -- cgit v1.2.3 From 180c4592e27ab808e1d85be113ef178ac95e7131 Mon Sep 17 00:00:00 2001 From: Wenjing Liu Date: Fri, 18 Mar 2022 12:22:35 -0400 Subject: drm/amd/display: add support for handling 128b/132b link training test request [why] DP2.x added new enum values for UHBR link rates in link training test request for test automation. We need to add UHBR link rates test request support in preparation for compliance test automation. [how] added a function that translate test link rate to dc link rate. Call the translation function to decide the requested test link rate. Tested-by: Daniel Wheeler Reviewed-by: George Shen Acked-by: Tom Chung Signed-off-by: Wenjing Liu Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c | 26 +++++++++++++++++++++++- drivers/gpu/drm/amd/display/dc/dc_dp_types.h | 12 ++++++++++- 2 files changed, 36 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c index 22dabe596dfc..c5f5d25035d2 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c @@ -4085,9 +4085,32 @@ static bool handle_hpd_irq_psr_sink(struct dc_link *link) return false; } +static enum dc_link_rate get_link_rate_from_test_link_rate(uint8_t test_rate) +{ + switch (test_rate) { + case DP_TEST_LINK_RATE_RBR: + return LINK_RATE_LOW; + case DP_TEST_LINK_RATE_HBR: + return LINK_RATE_HIGH; + case DP_TEST_LINK_RATE_HBR2: + return LINK_RATE_HIGH2; + case DP_TEST_LINK_RATE_HBR3: + return LINK_RATE_HIGH3; + case DP_TEST_LINK_RATE_UHBR10: + return LINK_RATE_UHBR10; + case DP_TEST_LINK_RATE_UHBR20: + return LINK_RATE_UHBR20; + case DP_TEST_LINK_RATE_UHBR13_5: + return LINK_RATE_UHBR13_5; + default: + return LINK_RATE_UNKNOWN; + } +} + static void dp_test_send_link_training(struct dc_link *link) { struct dc_link_settings link_settings = {0}; + uint8_t test_rate = 0; core_link_read_dpcd( link, @@ -4097,8 +4120,9 @@ static void dp_test_send_link_training(struct dc_link *link) core_link_read_dpcd( link, DP_TEST_LINK_RATE, - (unsigned char *)(&link_settings.link_rate), + &test_rate, 1); + link_settings.link_rate = get_link_rate_from_test_link_rate(test_rate); /* Set preferred link settings */ link->verified_link_cap.lane_count = link_settings.lane_count; diff --git a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h index 7d4aa99525da..2c54b6e0498b 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h @@ -133,6 +133,16 @@ enum dp_link_encoding { DP_128b_132b_ENCODING = 2, }; +enum dp_test_link_rate { + DP_TEST_LINK_RATE_RBR = 0x06, + DP_TEST_LINK_RATE_HBR = 0x0A, + DP_TEST_LINK_RATE_HBR2 = 0x14, + DP_TEST_LINK_RATE_HBR3 = 0x1E, + DP_TEST_LINK_RATE_UHBR10 = 0x01, + DP_TEST_LINK_RATE_UHBR20 = 0x02, + DP_TEST_LINK_RATE_UHBR13_5 = 0x03, +}; + struct dc_link_settings { enum dc_lane_count lane_count; enum dc_link_rate link_rate; @@ -620,7 +630,7 @@ union test_request { uint8_t LINK_TEST_PATTRN :1; uint8_t EDID_READ :1; uint8_t PHY_TEST_PATTERN :1; - uint8_t RESERVED :1; + uint8_t PHY_TEST_CHANNEL_CODING_TYPE :2; uint8_t AUDIO_TEST_PATTERN :1; uint8_t TEST_AUDIO_DISABLED_VIDEO :1; } bits; -- cgit v1.2.3 From 3f68934528633459e4bd7a986938612d65e13557 Mon Sep 17 00:00:00 2001 From: Martin Leung Date: Fri, 25 Mar 2022 12:48:57 -0400 Subject: drm/amd/display: revert Power down hardware if timer not trigger Tested-by: Daniel Wheeler Reviewed-by: George Shen Acked-by: Tom Chung Signed-off-by: Martin Leung Signed-off-by: Alex Deucher --- .../amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c | 26 +--------------------- 1 file changed, 1 insertion(+), 25 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c index 969b40250434..59fdd7f0d609 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c @@ -615,37 +615,13 @@ static void dcn31_clk_mgr_helper_populate_bw_params(struct clk_mgr_internal *clk } } -void dcn31_set_low_power_state(struct clk_mgr *clk_mgr_base) -{ - int display_count; - struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); - struct dc *dc = clk_mgr_base->ctx->dc; - struct dc_state *context = dc->current_state; - - if (clk_mgr_base->clks.pwr_state != DCN_PWR_STATE_LOW_POWER) { - display_count = dcn31_get_active_display_cnt_wa(dc, context); - /* if we can go lower, go lower */ - if (display_count == 0) { - union display_idle_optimization_u idle_info = { 0 }; - - idle_info.idle_info.df_request_disabled = 1; - idle_info.idle_info.phy_ref_clk_off = 1; - idle_info.idle_info.s0i2_rdy = 1; - dcn31_smu_set_display_idle_optimization(clk_mgr, idle_info.data); - /* update power state */ - clk_mgr_base->clks.pwr_state = DCN_PWR_STATE_LOW_POWER; - } - } -} - static struct clk_mgr_funcs dcn31_funcs = { .get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz, .update_clocks = dcn31_update_clocks, .init_clocks = dcn31_init_clocks, .enable_pme_wa = dcn31_enable_pme_wa, .are_clock_states_equal = dcn31_are_clock_states_equal, - .notify_wm_ranges = dcn31_notify_wm_ranges, - .set_low_power_state = dcn31_set_low_power_state + .notify_wm_ranges = dcn31_notify_wm_ranges }; extern struct clk_mgr_funcs dcn3_fpga_funcs; -- cgit v1.2.3 From eff5e11508244cba6b24e26d3673f06332c61cfc Mon Sep 17 00:00:00 2001 From: Wenjing Liu Date: Thu, 24 Mar 2022 17:24:55 -0400 Subject: drm/amd/display: extract set stream attribute to link_hwss [why] Extract set stream attribute into link_hwss as part of the link hwss refactor work. Tested-by: Daniel Wheeler Reviewed-by: George Shen Acked-by: Tom Chung Signed-off-by: Wenjing Liu Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc_link.c | 56 ++++------------------ drivers/gpu/drm/amd/display/dc/inc/link_hwss.h | 1 + .../gpu/drm/amd/display/dc/link/link_hwss_dio.c | 41 ++++++++++++++++ .../gpu/drm/amd/display/dc/link/link_hwss_dio.h | 1 + .../gpu/drm/amd/display/dc/link/link_hwss_dpia.c | 1 + .../gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c | 17 +++++++ .../drm/amd/display/dc/link/link_hwss_hpo_frl.c | 19 ++++++++ .../drm/amd/display/dc/virtual/virtual_link_hwss.c | 5 ++ 8 files changed, 94 insertions(+), 47 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index 7aede6495e5e..47b67fd1e84c 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -4123,6 +4123,7 @@ void core_link_enable_stream( struct link_encoder *link_enc; enum otg_out_mux_dest otg_out_dest = OUT_MUX_DIO; struct vpg *vpg = pipe_ctx->stream_res.stream_enc->vpg; + const struct link_hwss *link_hwss = get_link_hwss(link, &pipe_ctx->link_res); if (is_dp_128b_132b_signal(pipe_ctx)) vpg = pipe_ctx->stream_res.hpo_dp_stream_enc->vpg; @@ -4151,56 +4152,19 @@ void core_link_enable_stream( link_enc->funcs->setup( link_enc, pipe_ctx->stream->signal); - pipe_ctx->stream_res.stream_enc->funcs->setup_stereo_sync( - pipe_ctx->stream_res.stream_enc, - pipe_ctx->stream_res.tg->inst, - stream->timing.timing_3d_format != TIMING_3D_FORMAT_NONE); } - if (is_dp_128b_132b_signal(pipe_ctx)) { - pipe_ctx->stream_res.hpo_dp_stream_enc->funcs->set_stream_attribute( - pipe_ctx->stream_res.hpo_dp_stream_enc, - &stream->timing, - stream->output_color_space, - stream->use_vsc_sdp_for_colorimetry, - stream->timing.flags.DSC, - false); - otg_out_dest = OUT_MUX_HPO_DP; - } else if (dc_is_dp_signal(pipe_ctx->stream->signal)) { - pipe_ctx->stream_res.stream_enc->funcs->dp_set_stream_attribute( - pipe_ctx->stream_res.stream_enc, - &stream->timing, - stream->output_color_space, - stream->use_vsc_sdp_for_colorimetry, - stream->link->dpcd_caps.dprx_feature.bits.SST_SPLIT_SDP_CAP); - } - - if (dc_is_dp_signal(pipe_ctx->stream->signal)) - dp_source_sequence_trace(link, DPCD_SOURCE_SEQ_AFTER_DP_STREAM_ATTR); - - if (dc_is_hdmi_tmds_signal(pipe_ctx->stream->signal)) - pipe_ctx->stream_res.stream_enc->funcs->hdmi_set_stream_attribute( - pipe_ctx->stream_res.stream_enc, - &stream->timing, - stream->phy_pix_clk, - pipe_ctx->stream_res.audio != NULL); - pipe_ctx->stream->link->link_state_valid = true; - if (pipe_ctx->stream_res.tg->funcs->set_out_mux) + if (pipe_ctx->stream_res.tg->funcs->set_out_mux) { + if (is_dp_128b_132b_signal(pipe_ctx)) + otg_out_dest = OUT_MUX_HPO_DP; + else + otg_out_dest = OUT_MUX_DIO; pipe_ctx->stream_res.tg->funcs->set_out_mux(pipe_ctx->stream_res.tg, otg_out_dest); + } - if (dc_is_dvi_signal(pipe_ctx->stream->signal)) - pipe_ctx->stream_res.stream_enc->funcs->dvi_set_stream_attribute( - pipe_ctx->stream_res.stream_enc, - &stream->timing, - (pipe_ctx->stream->signal == SIGNAL_TYPE_DVI_DUAL_LINK) ? - true : false); - - if (dc_is_lvds_signal(pipe_ctx->stream->signal)) - pipe_ctx->stream_res.stream_enc->funcs->lvds_set_stream_attribute( - pipe_ctx->stream_res.stream_enc, - &stream->timing); + link_hwss->setup_stream_attribute(pipe_ctx); if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) { bool apply_edp_fast_boot_optimization = @@ -4335,13 +4299,11 @@ void core_link_enable_stream( dc->hwss.enable_audio_stream(pipe_ctx); } else { // if (IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) - if (is_dp_128b_132b_signal(pipe_ctx)) { + if (is_dp_128b_132b_signal(pipe_ctx)) fpga_dp_hpo_enable_link_and_stream(state, pipe_ctx); - } if (dc_is_dp_signal(pipe_ctx->stream->signal) || dc_is_virtual_signal(pipe_ctx->stream->signal)) dp_set_dsc_enable(pipe_ctx, true); - } if (pipe_ctx->stream->signal == SIGNAL_TYPE_HDMI_TYPE_A) { diff --git a/drivers/gpu/drm/amd/display/dc/inc/link_hwss.h b/drivers/gpu/drm/amd/display/dc/inc/link_hwss.h index 3b3090e3d327..e2f854464c05 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/link_hwss.h +++ b/drivers/gpu/drm/amd/display/dc/inc/link_hwss.h @@ -72,6 +72,7 @@ struct link_hwss { */ void (*setup_stream_encoder)(struct pipe_ctx *pipe_ctx); void (*reset_stream_encoder)(struct pipe_ctx *pipe_ctx); + void (*setup_stream_attribute)(struct pipe_ctx *pipe_ctx); }; #endif /* __DC_LINK_HWSS_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/link/link_hwss_dio.c b/drivers/gpu/drm/amd/display/dc/link/link_hwss_dio.c index 0f845113a6aa..e557c92acddc 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_hwss_dio.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_hwss_dio.c @@ -62,6 +62,46 @@ void reset_dio_stream_encoder(struct pipe_ctx *pipe_ctx) } +void setup_dio_stream_attribute(struct pipe_ctx *pipe_ctx) +{ + struct stream_encoder *stream_encoder = pipe_ctx->stream_res.stream_enc; + struct dc_stream_state *stream = pipe_ctx->stream; + struct dc_link *link = stream->link; + + if (!dc_is_virtual_signal(stream->signal)) + stream_encoder->funcs->setup_stereo_sync( + stream_encoder, + pipe_ctx->stream_res.tg->inst, + stream->timing.timing_3d_format != TIMING_3D_FORMAT_NONE); + + if (dc_is_dp_signal(stream->signal)) + stream_encoder->funcs->dp_set_stream_attribute( + stream_encoder, + &stream->timing, + stream->output_color_space, + stream->use_vsc_sdp_for_colorimetry, + link->dpcd_caps.dprx_feature.bits.SST_SPLIT_SDP_CAP); + else if (dc_is_hdmi_tmds_signal(stream->signal)) + stream_encoder->funcs->hdmi_set_stream_attribute( + stream_encoder, + &stream->timing, + stream->phy_pix_clk, + pipe_ctx->stream_res.audio != NULL); + else if (dc_is_dvi_signal(stream->signal)) + stream_encoder->funcs->dvi_set_stream_attribute( + stream_encoder, + &stream->timing, + (stream->signal == SIGNAL_TYPE_DVI_DUAL_LINK) ? + true : false); + else if (dc_is_lvds_signal(stream->signal)) + stream_encoder->funcs->lvds_set_stream_attribute( + stream_encoder, + &stream->timing); + + if (dc_is_dp_signal(stream->signal)) + dp_source_sequence_trace(link, DPCD_SOURCE_SEQ_AFTER_DP_STREAM_ATTR); +} + void enable_dio_dp_link_output(struct dc_link *link, const struct link_resource *link_res, enum signal_type signal, @@ -116,6 +156,7 @@ void set_dio_dp_lane_settings(struct dc_link *link, static const struct link_hwss dio_link_hwss = { .setup_stream_encoder = setup_dio_stream_encoder, .reset_stream_encoder = reset_dio_stream_encoder, + .setup_stream_attribute = setup_dio_stream_attribute, .ext = { .set_throttled_vcp_size = set_dio_throttled_vcp_size, .enable_dp_link_output = enable_dio_dp_link_output, diff --git a/drivers/gpu/drm/amd/display/dc/link/link_hwss_dio.h b/drivers/gpu/drm/amd/display/dc/link/link_hwss_dio.h index 680df20b1fa3..08f22b32df48 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_hwss_dio.h +++ b/drivers/gpu/drm/amd/display/dc/link/link_hwss_dio.h @@ -34,6 +34,7 @@ void set_dio_throttled_vcp_size(struct pipe_ctx *pipe_ctx, struct fixed31_32 throttled_vcp_size); void setup_dio_stream_encoder(struct pipe_ctx *pipe_ctx); void reset_dio_stream_encoder(struct pipe_ctx *pipe_ctx); +void setup_dio_stream_attribute(struct pipe_ctx *pipe_ctx); void enable_dio_dp_link_output(struct dc_link *link, const struct link_resource *link_res, enum signal_type signal, diff --git a/drivers/gpu/drm/amd/display/dc/link/link_hwss_dpia.c b/drivers/gpu/drm/amd/display/dc/link/link_hwss_dpia.c index 35b206225201..f90596f37cd4 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_hwss_dpia.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_hwss_dpia.c @@ -29,6 +29,7 @@ static const struct link_hwss dpia_link_hwss = { .setup_stream_encoder = setup_dio_stream_encoder, .reset_stream_encoder = reset_dio_stream_encoder, + .setup_stream_attribute = setup_dio_stream_attribute, .ext = { .set_throttled_vcp_size = set_dio_throttled_vcp_size, .enable_dp_link_output = enable_dio_dp_link_output, diff --git a/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c b/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c index 74919491675f..311e2e2687f0 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c @@ -131,6 +131,22 @@ static void reset_hpo_dp_stream_encoder(struct pipe_ctx *pipe_ctx) dccg->funcs->set_dpstreamclk(dccg, REFCLK, tg->inst); } +static void setup_hpo_dp_stream_attribute(struct pipe_ctx *pipe_ctx) +{ + struct hpo_dp_stream_encoder *stream_enc = pipe_ctx->stream_res.hpo_dp_stream_enc; + struct dc_stream_state *stream = pipe_ctx->stream; + struct dc_link *link = stream->link; + + stream_enc->funcs->set_stream_attribute( + stream_enc, + &stream->timing, + stream->output_color_space, + stream->use_vsc_sdp_for_colorimetry, + stream->timing.flags.DSC, + false); + dp_source_sequence_trace(link, DPCD_SOURCE_SEQ_AFTER_DP_STREAM_ATTR); +} + static void enable_hpo_dp_fpga_link_output(struct dc_link *link, const struct link_resource *link_res, enum signal_type signal, @@ -231,6 +247,7 @@ static void set_hpo_dp_lane_settings(struct dc_link *link, static const struct link_hwss hpo_dp_link_hwss = { .setup_stream_encoder = setup_hpo_dp_stream_encoder, .reset_stream_encoder = reset_hpo_dp_stream_encoder, + .setup_stream_attribute = setup_hpo_dp_stream_attribute, .ext = { .set_throttled_vcp_size = set_hpo_dp_throttled_vcp_size, .set_hblank_min_symbol_width = set_hpo_dp_hblank_min_symbol_width, diff --git a/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_frl.c b/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_frl.c index 9df273ca699b..4b5eccd994c4 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_frl.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_frl.c @@ -26,9 +26,28 @@ #include "core_types.h" #include "virtual/virtual_link_hwss.h" +static void setup_hpo_frl_stream_attribute(struct pipe_ctx *pipe_ctx) +{ + struct hpo_frl_stream_encoder *stream_enc = pipe_ctx->stream_res.hpo_frl_stream_enc; + struct dc_stream_state *stream = pipe_ctx->stream; + struct pipe_ctx *odm_pipe; + int odm_combine_num_segments = 1; + + /* get number of ODM combine input segments */ + for (odm_pipe = pipe_ctx->next_odm_pipe; odm_pipe; odm_pipe = odm_pipe->next_odm_pipe) + odm_combine_num_segments++; + + stream_enc->funcs->hdmi_frl_set_stream_attribute( + stream_enc, + &stream->timing, + &stream->link->frl_link_settings.borrow_params, + odm_combine_num_segments); +} + static const struct link_hwss hpo_frl_link_hwss = { .setup_stream_encoder = virtual_setup_stream_encoder, .reset_stream_encoder = virtual_reset_stream_encoder, + .setup_stream_attribute = setup_hpo_frl_stream_attribute, }; bool can_use_hpo_frl_link_hwss(const struct dc_link *link, diff --git a/drivers/gpu/drm/amd/display/dc/virtual/virtual_link_hwss.c b/drivers/gpu/drm/amd/display/dc/virtual/virtual_link_hwss.c index 525eba2a3354..501173ce270e 100644 --- a/drivers/gpu/drm/amd/display/dc/virtual/virtual_link_hwss.c +++ b/drivers/gpu/drm/amd/display/dc/virtual/virtual_link_hwss.c @@ -29,12 +29,17 @@ void virtual_setup_stream_encoder(struct pipe_ctx *pipe_ctx) { } +void virtual_setup_stream_attribute(struct pipe_ctx *pipe_ctx) +{ +} + void virtual_reset_stream_encoder(struct pipe_ctx *pipe_ctx) { } static const struct link_hwss virtual_link_hwss = { .setup_stream_encoder = virtual_setup_stream_encoder, .reset_stream_encoder = virtual_reset_stream_encoder, + .setup_stream_attribute = virtual_setup_stream_attribute, }; const struct link_hwss *get_virtual_link_hwss(void) -- cgit v1.2.3 From 0191d384176c00fa28b52393997c61fce346564f Mon Sep 17 00:00:00 2001 From: Krunoslav Kovac Date: Fri, 25 Mar 2022 14:41:49 -0400 Subject: drm/amd/display: Enable 3 plane for DCN 3.0 and 3.02 [WHY&HOW] Increase num of bottom planes to 2. Note that DCN 3.03 is left out since it has max 2 planes. Tested-by: Daniel Wheeler Reviewed-by: Aric Cyr Acked-by: Tom Chung Signed-off-by: Krunoslav Kovac Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c | 6 +++--- drivers/gpu/drm/amd/display/dc/dcn302/dcn302_resource.c | 6 +++--- 2 files changed, 6 insertions(+), 6 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c index e6a62cc75139..336b2ce6a636 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_resource.c @@ -2602,9 +2602,9 @@ static bool dcn30_resource_construct( dc->caps.mall_size_total = dc->caps.mall_size_per_mem_channel * dc->ctx->dc_bios->vram_info.num_chans * 1048576; dc->caps.cursor_cache_size = dc->caps.max_cursor_size * dc->caps.max_cursor_size * 8; - dc->caps.max_slave_planes = 1; - dc->caps.max_slave_yuv_planes = 1; - dc->caps.max_slave_rgb_planes = 1; + dc->caps.max_slave_planes = 2; + dc->caps.max_slave_yuv_planes = 2; + dc->caps.max_slave_rgb_planes = 2; dc->caps.post_blend_color_processing = true; dc->caps.force_dp_tps4_for_cp2520 = true; dc->caps.extended_aux_timeout_support = true; diff --git a/drivers/gpu/drm/amd/display/dc/dcn302/dcn302_resource.c b/drivers/gpu/drm/amd/display/dc/dcn302/dcn302_resource.c index 88318e8ffca8..f0938653bb88 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn302/dcn302_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn302/dcn302_resource.c @@ -1219,9 +1219,9 @@ static bool dcn302_resource_construct( /* total size = mall per channel * num channels * 1024 * 1024 */ dc->caps.mall_size_total = dc->caps.mall_size_per_mem_channel * dc->ctx->dc_bios->vram_info.num_chans * 1048576; dc->caps.cursor_cache_size = dc->caps.max_cursor_size * dc->caps.max_cursor_size * 8; - dc->caps.max_slave_planes = 1; - dc->caps.max_slave_yuv_planes = 1; - dc->caps.max_slave_rgb_planes = 1; + dc->caps.max_slave_planes = 2; + dc->caps.max_slave_yuv_planes = 2; + dc->caps.max_slave_rgb_planes = 2; dc->caps.post_blend_color_processing = true; dc->caps.force_dp_tps4_for_cp2520 = true; dc->caps.extended_aux_timeout_support = true; -- cgit v1.2.3 From 24566aab61c98116e6c39db7b27b005f6d7776d8 Mon Sep 17 00:00:00 2001 From: Becle Lee Date: Mon, 28 Mar 2022 11:18:23 +0800 Subject: drm/amd/display: fix missing-prototypes warning [Why] No declaration of hubp1_wait_pipe_read_start found in header file. [How] Add its declaration. Tested-by: Daniel Wheeler Reviewed-by: Aric Cyr Acked-by: Tom Chung Signed-off-by: Becle Lee Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h | 1 + 1 file changed, 1 insertion(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h index 0b17c2993ca5..4ccb03a7564a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h @@ -790,5 +790,6 @@ bool hubp1_in_blank(struct hubp *hubp); void hubp1_soft_reset(struct hubp *hubp, bool reset); void hubp1_set_flip_int(struct hubp *hubp); +void hubp1_wait_pipe_read_start(struct hubp *hubp); #endif -- cgit v1.2.3 From 68cdbf631b56a97c75891d8292900791713ba0ad Mon Sep 17 00:00:00 2001 From: Aric Cyr Date: Mon, 28 Mar 2022 09:41:01 -0400 Subject: drm/amd/display: 3.2.180 - [FW Promotion] Release 0.0.111.0 - Check for invalid input params when building scaling params - Move link_trace for edp to dp_trace - Fix missing-prototypes warning - Enable 3 plane for DCN 3.0 and 3.02 - Extract set stream attribute into link_hwss - Revert Power down hardware if timer not trigger - Add support for handling 128b/132b link training test request - Add configuration options for AUX wake work around - Remove underflow IRQ type - Add flip interval workaround for low FPS in some game - Remove assert for odm transition case Tested-by: Daniel Wheeler Acked-by: Tom Chung Signed-off-by: Aric Cyr Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dc.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index e148db770722..70cc21ec643a 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -47,7 +47,7 @@ struct aux_payload; struct set_config_cmd_payload; struct dmub_notification; -#define DC_VER "3.2.179" +#define DC_VER "3.2.180" #define MAX_SURFACES 3 #define MAX_PLANES 6 -- cgit v1.2.3 From 176cd4385aea1e1c1e41c1b8e02764558e4f6a3c Mon Sep 17 00:00:00 2001 From: "Leo (Hanghong) Ma" Date: Wed, 23 Mar 2022 15:35:57 -0400 Subject: drm/amd/display: Move link_trace for edp to dp_trace [Why & How] The dp_trace structure is self contained component designed for all dp trace, and the edp link trace should be a part of it; Suggested-by: Wenjing Liu Tested-by: Daniel Wheeler Reviewed-by: Wenjing Liu Reviewed-by: Charlene Liu Acked-by: Tom Chung Signed-off-by: Leo (Hanghong) Ma Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dc_link.h | 12 ++++------- .../amd/display/dc/dce110/dce110_hw_sequencer.c | 25 ++++++++++------------ .../gpu/drm/amd/display/dc/link/link_dp_trace.c | 20 +++++++++++++++++ .../gpu/drm/amd/display/dc/link/link_dp_trace.h | 5 +++++ 4 files changed, 40 insertions(+), 22 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dc_link.h b/drivers/gpu/drm/amd/display/dc/dc_link.h index aa818bf840eb..a3c37ee3f849 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_link.h +++ b/drivers/gpu/drm/amd/display/dc/dc_link.h @@ -67,13 +67,9 @@ struct link_mst_stream_allocation_table { struct link_mst_stream_allocation stream_allocations[MAX_CONTROLLER_NUM]; }; -struct time_stamp { - uint64_t edp_poweroff; - uint64_t edp_poweron; -}; - -struct link_trace { - struct time_stamp time_stamp; +struct edp_trace_power_timestamps { + uint64_t poweroff; + uint64_t poweron; }; struct dp_trace_lt_counts { @@ -96,6 +92,7 @@ struct dp_trace { struct dp_trace_lt commit_lt_trace; unsigned int link_loss_count; bool is_initialized; + struct edp_trace_power_timestamps edp_trace_power_timestamps; }; /* PSR feature flags */ @@ -231,7 +228,6 @@ struct dc_link { struct dc_link_status link_status; struct dprx_states dprx_states; - struct link_trace link_trace; struct gpio *hpd_gpio; enum dc_link_fec_state fec_state; }; diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c index 248602c15f3a..9fc1ba12ec19 100644 --- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c @@ -67,6 +67,7 @@ #include "dcn10/dcn10_hw_sequencer.h" +#include "link/link_dp_trace.h" #include "dce110_hw_sequencer.h" #define GAMMA_HW_POINTS_NUM 256 @@ -819,19 +820,19 @@ void dce110_edp_power_control( div64_u64(dm_get_elapse_time_in_ns( ctx, current_ts, - link->link_trace.time_stamp.edp_poweroff), 1000000); + dp_trace_get_edp_poweroff_timestamp(link)), 1000000); unsigned long long time_since_edp_poweron_ms = div64_u64(dm_get_elapse_time_in_ns( ctx, current_ts, - link->link_trace.time_stamp.edp_poweron), 1000000); + dp_trace_get_edp_poweron_timestamp(link)), 1000000); DC_LOG_HW_RESUME_S3( "%s: transition: power_up=%d current_ts=%llu edp_poweroff=%llu edp_poweron=%llu time_since_edp_poweroff_ms=%llu time_since_edp_poweron_ms=%llu", __func__, power_up, current_ts, - link->link_trace.time_stamp.edp_poweroff, - link->link_trace.time_stamp.edp_poweron, + dp_trace_get_edp_poweroff_timestamp(link), + dp_trace_get_edp_poweron_timestamp(link), time_since_edp_poweroff_ms, time_since_edp_poweron_ms); @@ -846,7 +847,7 @@ void dce110_edp_power_control( link->local_sink->edid_caps.panel_patch.extra_t12_ms; /* Adjust remaining_min_edp_poweroff_time_ms if this is not the first time. */ - if (link->link_trace.time_stamp.edp_poweroff != 0) { + if (dp_trace_get_edp_poweroff_timestamp(link) != 0) { if (time_since_edp_poweroff_ms < remaining_min_edp_poweroff_time_ms) remaining_min_edp_poweroff_time_ms = remaining_min_edp_poweroff_time_ms - time_since_edp_poweroff_ms; @@ -904,17 +905,13 @@ void dce110_edp_power_control( __func__, (power_up ? "On":"Off"), bp_result); - if (!power_up) - /*save driver power off time stamp*/ - link->link_trace.time_stamp.edp_poweroff = dm_get_timestamp(ctx); - else - link->link_trace.time_stamp.edp_poweron = dm_get_timestamp(ctx); + dp_trace_set_edp_power_timestamp(link, power_up); DC_LOG_HW_RESUME_S3( "%s: updated values: edp_poweroff=%llu edp_poweron=%llu\n", __func__, - link->link_trace.time_stamp.edp_poweroff, - link->link_trace.time_stamp.edp_poweron); + dp_trace_get_edp_poweroff_timestamp(link), + dp_trace_get_edp_poweron_timestamp(link)); if (bp_result != BP_RESULT_OK) DC_LOG_ERROR( @@ -942,14 +939,14 @@ void dce110_edp_wait_for_T12( return; if (!link->panel_cntl->funcs->is_panel_powered_on(link->panel_cntl) && - link->link_trace.time_stamp.edp_poweroff != 0) { + dp_trace_get_edp_poweroff_timestamp(link) != 0) { unsigned int t12_duration = 500; // Default T12 as per spec unsigned long long current_ts = dm_get_timestamp(ctx); unsigned long long time_since_edp_poweroff_ms = div64_u64(dm_get_elapse_time_in_ns( ctx, current_ts, - link->link_trace.time_stamp.edp_poweroff), 1000000); + dp_trace_get_edp_poweroff_timestamp(link)), 1000000); t12_duration += link->local_sink->edid_caps.panel_patch.extra_t12_ms; // Add extra T12 diff --git a/drivers/gpu/drm/amd/display/dc/link/link_dp_trace.c b/drivers/gpu/drm/amd/display/dc/link/link_dp_trace.c index e7047391934b..2c1a3bfcdb50 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_dp_trace.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_dp_trace.c @@ -144,3 +144,23 @@ unsigned int dc_dp_trace_get_link_loss_count(struct dc_link *link) { return link->dp_trace.link_loss_count; } + +void dp_trace_set_edp_power_timestamp(struct dc_link *link, + bool power_up) +{ + if (!power_up) + /*save driver power off time stamp*/ + link->dp_trace.edp_trace_power_timestamps.poweroff = dm_get_timestamp(link->dc->ctx); + else + link->dp_trace.edp_trace_power_timestamps.poweron = dm_get_timestamp(link->dc->ctx); +} + +uint64_t dp_trace_get_edp_poweron_timestamp(struct dc_link *link) +{ + return link->dp_trace.edp_trace_power_timestamps.poweron; +} + +uint64_t dp_trace_get_edp_poweroff_timestamp(struct dc_link *link) +{ + return link->dp_trace.edp_trace_power_timestamps.poweroff; +} \ No newline at end of file diff --git a/drivers/gpu/drm/amd/display/dc/link/link_dp_trace.h b/drivers/gpu/drm/amd/display/dc/link/link_dp_trace.h index 702f97c6ead0..26700e3cd65e 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_dp_trace.h +++ b/drivers/gpu/drm/amd/display/dc/link/link_dp_trace.h @@ -54,4 +54,9 @@ struct dp_trace_lt_counts *dc_dp_trace_get_lt_counts(struct dc_link *link, bool in_detection); unsigned int dc_dp_trace_get_link_loss_count(struct dc_link *link); +void dp_trace_set_edp_power_timestamp(struct dc_link *link, + bool power_up); +uint64_t dp_trace_get_edp_poweron_timestamp(struct dc_link *link); +uint64_t dp_trace_get_edp_poweroff_timestamp(struct dc_link *link); + #endif /* __LINK_DP_TRACE_H__ */ -- cgit v1.2.3 From 73b1da69f5314d96e1f963468863eaa884ee1030 Mon Sep 17 00:00:00 2001 From: Michael Strauss Date: Wed, 23 Mar 2022 16:06:25 -0400 Subject: drm/amd/display: Check for invalid input params when building scaling params [WHY] Function to calculate scaling ratios can be called with invalid plane src/dest, causing a divide by zero. [HOW] Fail building scaling params if plane state src/dest rects are unpopulated Tested-by: Daniel Wheeler Reviewed-by: Charlene Liu Acked-by: Tom Chung Signed-off-by: Michael Strauss Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c index d251c3f3a714..f5777a71f2f1 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c @@ -1076,6 +1076,15 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx) bool res = false; DC_LOGGER_INIT(pipe_ctx->stream->ctx->logger); + /* Invalid input */ + if (!plane_state->dst_rect.width || + !plane_state->dst_rect.height || + !plane_state->src_rect.width || + !plane_state->src_rect.height) { + ASSERT(0); + return false; + } + pipe_ctx->plane_res.scl_data.format = convert_pixel_format_to_dalsurface( pipe_ctx->plane_state->format); -- cgit v1.2.3 From 78174f477081b61b21c6f8b616234b00ce556840 Mon Sep 17 00:00:00 2001 From: Anthony Koo Date: Sat, 26 Mar 2022 15:20:21 -0400 Subject: drm/amd/display: [FW Promotion] Release 0.0.111.0 - Add options to allow for configurable PHY options during PSR active state - Remove unused versioning and git hash Tested-by: Daniel Wheeler Reviewed-by: Aric Cyr Acked-by: Tom Chung Signed-off-by: Anthony Koo Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h | 102 +++++++++++++++++++----- 1 file changed, 82 insertions(+), 20 deletions(-) (limited to 'drivers/gpu') 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 95f1c7198dd3..05c8d91ad4ab 100644 --- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h @@ -44,24 +44,6 @@ #endif // defined(_TEST_HARNESS) || defined(FPGA_USB4) -/* Firmware versioning. */ -#ifdef DMUB_EXPOSE_VERSION -#define DMUB_FW_VERSION_GIT_HASH 0x19edd13d -#define DMUB_FW_VERSION_MAJOR 0 -#define DMUB_FW_VERSION_MINOR 0 -#define DMUB_FW_VERSION_REVISION 110 -#define DMUB_FW_VERSION_TEST 0 -#define DMUB_FW_VERSION_VBIOS 0 -#define DMUB_FW_VERSION_HOTFIX 0 -#define DMUB_FW_VERSION_UCODE (((DMUB_FW_VERSION_MAJOR & 0xFF) << 24) | \ - ((DMUB_FW_VERSION_MINOR & 0xFF) << 16) | \ - ((DMUB_FW_VERSION_REVISION & 0xFF) << 8) | \ - ((DMUB_FW_VERSION_TEST & 0x1) << 7) | \ - ((DMUB_FW_VERSION_VBIOS & 0x1) << 6) | \ - (DMUB_FW_VERSION_HOTFIX & 0x3F)) - -#endif - //================================================================== /* Basic type definitions. */ @@ -1450,6 +1432,79 @@ enum dmub_cmd_mall_type { DMUB_CMD__MALL_ACTION_NO_DF_REQ = 3, }; +/** + * PHY Link rate for DP. + */ +enum phy_link_rate { + /** + * not supported. + */ + PHY_RATE_UNKNOWN = 0, + /** + * Rate_1 (RBR) - 1.62 Gbps/Lane + */ + PHY_RATE_162 = 1, + /** + * Rate_2 - 2.16 Gbps/Lane + */ + PHY_RATE_216 = 2, + /** + * Rate_3 - 2.43 Gbps/Lane + */ + PHY_RATE_243 = 3, + /** + * Rate_4 (HBR) - 2.70 Gbps/Lane + */ + PHY_RATE_270 = 4, + /** + * Rate_5 (RBR2)- 3.24 Gbps/Lane + */ + PHY_RATE_324 = 5, + /** + * Rate_6 - 4.32 Gbps/Lane + */ + PHY_RATE_432 = 6, + /** + * Rate_7 (HBR2)- 5.40 Gbps/Lane + */ + PHY_RATE_540 = 7, + /** + * Rate_8 (HBR3)- 8.10 Gbps/Lane + */ + PHY_RATE_810 = 8, + /** + * UHBR10 - 10.0 Gbps/Lane + */ + PHY_RATE_1000 = 9, + /** + * UHBR13.5 - 13.5 Gbps/Lane + */ + PHY_RATE_1350 = 10, + /** + * UHBR10 - 20.0 Gbps/Lane + */ + PHY_RATE_2000 = 11, +}; + +/** + * enum dmub_phy_fsm_state - PHY FSM states. + * PHY FSM state to transit to during PSR enable/disable. + */ +enum dmub_phy_fsm_state { + DMUB_PHY_FSM_POWER_UP_DEFAULT = 0, + DMUB_PHY_FSM_RESET, + DMUB_PHY_FSM_RESET_RELEASED, + DMUB_PHY_FSM_SRAM_LOAD_DONE, + DMUB_PHY_FSM_INITIALIZED, + DMUB_PHY_FSM_CALIBRATED, + DMUB_PHY_FSM_CALIBRATED_LP, + DMUB_PHY_FSM_CALIBRATED_PG, + DMUB_PHY_FSM_POWER_DOWN, + DMUB_PHY_FSM_PLL_EN, + DMUB_PHY_FSM_TX_EN, + DMUB_PHY_FSM_FAST_LP, +}; + /** * Data passed from driver to FW in a DMUB_CMD__PSR_COPY_SETTINGS command. */ @@ -1629,9 +1684,16 @@ struct dmub_rb_cmd_psr_enable_data { */ uint8_t panel_inst; /** - * Explicit padding to 4 byte boundary. + * Phy state to enter. + * Values to use are defined in dmub_phy_fsm_state */ - uint8_t pad[2]; + uint8_t phy_fsm_state; + /** + * Phy rate for DP - RBR/HBR/HBR2/HBR3. + * Set this using enum phy_link_rate. + * This does not support HDMI/DP2 for now. + */ + uint8_t phy_rate; }; /** -- cgit v1.2.3 From 96621ca578dbadeb12bd190e0733cdc5c76899d8 Mon Sep 17 00:00:00 2001 From: Philip Yang Date: Mon, 4 Apr 2022 17:25:23 -0400 Subject: drm/amdkfd: Add missing NULL check in svm_range_map_to_gpu MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit bo_adev is NULL for system memory mapping to GPU. Fixes: 30671b44aa570a ("drm/amdgpu: fix TLB flushing during eviction") Signed-off-by: Philip Yang Reviewed-by: Felix Kuehling Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index 907b02045824..d3fb2d0b5a25 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -1281,7 +1281,7 @@ svm_range_map_to_gpu(struct kfd_process_device *pdd, struct svm_range *prange, last_start, prange->start + i, pte_flags, last_start - prange->start, - bo_adev->vm_manager.vram_base_offset, + bo_adev ? bo_adev->vm_manager.vram_base_offset : 0, NULL, dma_addr, &vm->last_update); for (j = last_start - prange->start; j <= i; j++) -- cgit v1.2.3 From 69f0b547e68c27a2e432a2d0459915babf015cb3 Mon Sep 17 00:00:00 2001 From: Guo Zhengkui Date: Fri, 1 Apr 2022 10:59:29 +0800 Subject: drm/radeon: simplify if-if to if-else Replace `if (!ret)` with `else` for simplification. Signed-off-by: Guo Zhengkui Signed-off-by: Alex Deucher --- drivers/gpu/drm/radeon/radeon_pm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c index c67b6ddb29a4..e765abcb3b01 100644 --- a/drivers/gpu/drm/radeon/radeon_pm.c +++ b/drivers/gpu/drm/radeon/radeon_pm.c @@ -1629,7 +1629,7 @@ int radeon_pm_late_init(struct radeon_device *rdev) ret = device_create_file(rdev->dev, &dev_attr_power_method); if (ret) DRM_ERROR("failed to create device file for power method\n"); - if (!ret) + else rdev->pm.sysfs_initialized = true; } } -- cgit v1.2.3 From fa458eb10dc7218146a84e6d2e072424e64d188a Mon Sep 17 00:00:00 2001 From: Benjamin Marty Date: Wed, 23 Mar 2022 22:08:26 +0100 Subject: drm/amdgpu/display: change pipe policy for DCN 2.1 Fixes crash on MST Hub disconnect. Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1849 Fixes: ee2698cf79cc ("drm/amd/display: Changed pipe split policy to allow for multi-display pipe split") Signed-off-by: Benjamin Marty Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c index 612732656772..3fe4bfbb98a0 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c @@ -644,7 +644,7 @@ static const struct dc_debug_options debug_defaults_drv = { .clock_trace = true, .disable_pplib_clock_request = true, .min_disp_clk_khz = 100000, - .pipe_split_policy = MPC_SPLIT_DYNAMIC, + .pipe_split_policy = MPC_SPLIT_AVOID_MULT_DISP, .force_single_disp_pipe_split = false, .disable_dcc = DCC_ENABLE, .vsr_support = true, -- cgit v1.2.3 From b0778bb0afb10d36de47a435abdc4ca4ecfc3f89 Mon Sep 17 00:00:00 2001 From: Tom Rix Date: Mon, 4 Apr 2022 18:57:10 -0400 Subject: drm/radeon: change si_default_state table from global to static MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Smatch reports these issues si_blit_shaders.c:31:11: warning: symbol 'si_default_state' was not declared. Should it be static? si_blit_shaders.c:253:11: warning: symbol 'si_default_size' was not declared. Should it be static? Both symbols are only used in si.c. Single file symbols should be static. So move the definition of si_default_state and si_default_size to si_blit_shader.h and change their storage-class-specifier to static. Remove unneeded si_blit_shader.c Reviewed-by: Christian König Signed-off-by: Tom Rix Signed-off-by: Alex Deucher --- drivers/gpu/drm/radeon/Makefile | 2 +- drivers/gpu/drm/radeon/si_blit_shaders.c | 253 ------------------------------- drivers/gpu/drm/radeon/si_blit_shaders.h | 223 ++++++++++++++++++++++++++- 3 files changed, 222 insertions(+), 256 deletions(-) delete mode 100644 drivers/gpu/drm/radeon/si_blit_shaders.c (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile index 11c97edde54d..664381f4eb07 100644 --- a/drivers/gpu/drm/radeon/Makefile +++ b/drivers/gpu/drm/radeon/Makefile @@ -44,7 +44,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \ evergreen.o evergreen_cs.o evergreen_blit_shaders.o \ evergreen_hdmi.o radeon_trace_points.o ni.o cayman_blit_shaders.o \ atombios_encoders.o radeon_semaphore.o radeon_sa.o atombios_i2c.o si.o \ - si_blit_shaders.o radeon_prime.o cik.o cik_blit_shaders.o \ + radeon_prime.o cik.o cik_blit_shaders.o \ r600_dpm.o rs780_dpm.o rv6xx_dpm.o rv770_dpm.o rv730_dpm.o rv740_dpm.o \ rv770_smc.o cypress_dpm.o btc_dpm.o sumo_dpm.o sumo_smc.o trinity_dpm.o \ trinity_smc.o ni_dpm.o si_smc.o si_dpm.o kv_smc.o kv_dpm.o ci_smc.o \ diff --git a/drivers/gpu/drm/radeon/si_blit_shaders.c b/drivers/gpu/drm/radeon/si_blit_shaders.c deleted file mode 100644 index ec415e7dfa4b..000000000000 --- a/drivers/gpu/drm/radeon/si_blit_shaders.c +++ /dev/null @@ -1,253 +0,0 @@ -/* - * Copyright 2011 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice (including the next - * paragraph) shall be included in all copies or substantial portions of the - * Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER - * DEALINGS IN THE SOFTWARE. - * - * Authors: - * Alex Deucher - */ - -#include -#include -#include - -const u32 si_default_state[] = -{ - 0xc0066900, - 0x00000000, - 0x00000060, /* DB_RENDER_CONTROL */ - 0x00000000, /* DB_COUNT_CONTROL */ - 0x00000000, /* DB_DEPTH_VIEW */ - 0x0000002a, /* DB_RENDER_OVERRIDE */ - 0x00000000, /* DB_RENDER_OVERRIDE2 */ - 0x00000000, /* DB_HTILE_DATA_BASE */ - - 0xc0046900, - 0x00000008, - 0x00000000, /* DB_DEPTH_BOUNDS_MIN */ - 0x00000000, /* DB_DEPTH_BOUNDS_MAX */ - 0x00000000, /* DB_STENCIL_CLEAR */ - 0x00000000, /* DB_DEPTH_CLEAR */ - - 0xc0036900, - 0x0000000f, - 0x00000000, /* DB_DEPTH_INFO */ - 0x00000000, /* DB_Z_INFO */ - 0x00000000, /* DB_STENCIL_INFO */ - - 0xc0016900, - 0x00000080, - 0x00000000, /* PA_SC_WINDOW_OFFSET */ - - 0xc00d6900, - 0x00000083, - 0x0000ffff, /* PA_SC_CLIPRECT_RULE */ - 0x00000000, /* PA_SC_CLIPRECT_0_TL */ - 0x20002000, /* PA_SC_CLIPRECT_0_BR */ - 0x00000000, - 0x20002000, - 0x00000000, - 0x20002000, - 0x00000000, - 0x20002000, - 0xaaaaaaaa, /* PA_SC_EDGERULE */ - 0x00000000, /* PA_SU_HARDWARE_SCREEN_OFFSET */ - 0x0000000f, /* CB_TARGET_MASK */ - 0x0000000f, /* CB_SHADER_MASK */ - - 0xc0226900, - 0x00000094, - 0x80000000, /* PA_SC_VPORT_SCISSOR_0_TL */ - 0x20002000, /* PA_SC_VPORT_SCISSOR_0_BR */ - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x00000000, /* PA_SC_VPORT_ZMIN_0 */ - 0x3f800000, /* PA_SC_VPORT_ZMAX_0 */ - - 0xc0026900, - 0x000000d9, - 0x00000000, /* CP_RINGID */ - 0x00000000, /* CP_VMID */ - - 0xc0046900, - 0x00000100, - 0xffffffff, /* VGT_MAX_VTX_INDX */ - 0x00000000, /* VGT_MIN_VTX_INDX */ - 0x00000000, /* VGT_INDX_OFFSET */ - 0x00000000, /* VGT_MULTI_PRIM_IB_RESET_INDX */ - - 0xc0046900, - 0x00000105, - 0x00000000, /* CB_BLEND_RED */ - 0x00000000, /* CB_BLEND_GREEN */ - 0x00000000, /* CB_BLEND_BLUE */ - 0x00000000, /* CB_BLEND_ALPHA */ - - 0xc0016900, - 0x000001e0, - 0x00000000, /* CB_BLEND0_CONTROL */ - - 0xc00e6900, - 0x00000200, - 0x00000000, /* DB_DEPTH_CONTROL */ - 0x00000000, /* DB_EQAA */ - 0x00cc0010, /* CB_COLOR_CONTROL */ - 0x00000210, /* DB_SHADER_CONTROL */ - 0x00010000, /* PA_CL_CLIP_CNTL */ - 0x00000004, /* PA_SU_SC_MODE_CNTL */ - 0x00000100, /* PA_CL_VTE_CNTL */ - 0x00000000, /* PA_CL_VS_OUT_CNTL */ - 0x00000000, /* PA_CL_NANINF_CNTL */ - 0x00000000, /* PA_SU_LINE_STIPPLE_CNTL */ - 0x00000000, /* PA_SU_LINE_STIPPLE_SCALE */ - 0x00000000, /* PA_SU_PRIM_FILTER_CNTL */ - 0x00000000, /* */ - 0x00000000, /* */ - - 0xc0116900, - 0x00000280, - 0x00000000, /* PA_SU_POINT_SIZE */ - 0x00000000, /* PA_SU_POINT_MINMAX */ - 0x00000008, /* PA_SU_LINE_CNTL */ - 0x00000000, /* PA_SC_LINE_STIPPLE */ - 0x00000000, /* VGT_OUTPUT_PATH_CNTL */ - 0x00000000, /* VGT_HOS_CNTL */ - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, /* VGT_GS_MODE */ - - 0xc0026900, - 0x00000292, - 0x00000000, /* PA_SC_MODE_CNTL_0 */ - 0x00000000, /* PA_SC_MODE_CNTL_1 */ - - 0xc0016900, - 0x000002a1, - 0x00000000, /* VGT_PRIMITIVEID_EN */ - - 0xc0016900, - 0x000002a5, - 0x00000000, /* VGT_MULTI_PRIM_IB_RESET_EN */ - - 0xc0026900, - 0x000002a8, - 0x00000000, /* VGT_INSTANCE_STEP_RATE_0 */ - 0x00000000, - - 0xc0026900, - 0x000002ad, - 0x00000000, /* VGT_REUSE_OFF */ - 0x00000000, - - 0xc0016900, - 0x000002d5, - 0x00000000, /* VGT_SHADER_STAGES_EN */ - - 0xc0016900, - 0x000002dc, - 0x0000aa00, /* DB_ALPHA_TO_MASK */ - - 0xc0066900, - 0x000002de, - 0x00000000, /* PA_SU_POLY_OFFSET_DB_FMT_CNTL */ - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - - 0xc0026900, - 0x000002e5, - 0x00000000, /* VGT_STRMOUT_CONFIG */ - 0x00000000, - - 0xc01b6900, - 0x000002f5, - 0x76543210, /* PA_SC_CENTROID_PRIORITY_0 */ - 0xfedcba98, /* PA_SC_CENTROID_PRIORITY_1 */ - 0x00000000, /* PA_SC_LINE_CNTL */ - 0x00000000, /* PA_SC_AA_CONFIG */ - 0x00000005, /* PA_SU_VTX_CNTL */ - 0x3f800000, /* PA_CL_GB_VERT_CLIP_ADJ */ - 0x3f800000, /* PA_CL_GB_VERT_DISC_ADJ */ - 0x3f800000, /* PA_CL_GB_HORZ_CLIP_ADJ */ - 0x3f800000, /* PA_CL_GB_HORZ_DISC_ADJ */ - 0x00000000, /* PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_0 */ - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0xffffffff, /* PA_SC_AA_MASK_X0Y0_X1Y0 */ - 0xffffffff, - - 0xc0026900, - 0x00000316, - 0x0000000e, /* VGT_VERTEX_REUSE_BLOCK_CNTL */ - 0x00000010, /* */ -}; - -const u32 si_default_size = ARRAY_SIZE(si_default_state); diff --git a/drivers/gpu/drm/radeon/si_blit_shaders.h b/drivers/gpu/drm/radeon/si_blit_shaders.h index c739e51e3961..829a2b6228b7 100644 --- a/drivers/gpu/drm/radeon/si_blit_shaders.h +++ b/drivers/gpu/drm/radeon/si_blit_shaders.h @@ -25,8 +25,227 @@ #ifndef SI_BLIT_SHADERS_H #define SI_BLIT_SHADERS_H -extern const u32 si_default_state[]; +static const u32 si_default_state[] = { + 0xc0066900, + 0x00000000, + 0x00000060, /* DB_RENDER_CONTROL */ + 0x00000000, /* DB_COUNT_CONTROL */ + 0x00000000, /* DB_DEPTH_VIEW */ + 0x0000002a, /* DB_RENDER_OVERRIDE */ + 0x00000000, /* DB_RENDER_OVERRIDE2 */ + 0x00000000, /* DB_HTILE_DATA_BASE */ -extern const u32 si_default_size; + 0xc0046900, + 0x00000008, + 0x00000000, /* DB_DEPTH_BOUNDS_MIN */ + 0x00000000, /* DB_DEPTH_BOUNDS_MAX */ + 0x00000000, /* DB_STENCIL_CLEAR */ + 0x00000000, /* DB_DEPTH_CLEAR */ + + 0xc0036900, + 0x0000000f, + 0x00000000, /* DB_DEPTH_INFO */ + 0x00000000, /* DB_Z_INFO */ + 0x00000000, /* DB_STENCIL_INFO */ + + 0xc0016900, + 0x00000080, + 0x00000000, /* PA_SC_WINDOW_OFFSET */ + + 0xc00d6900, + 0x00000083, + 0x0000ffff, /* PA_SC_CLIPRECT_RULE */ + 0x00000000, /* PA_SC_CLIPRECT_0_TL */ + 0x20002000, /* PA_SC_CLIPRECT_0_BR */ + 0x00000000, + 0x20002000, + 0x00000000, + 0x20002000, + 0x00000000, + 0x20002000, + 0xaaaaaaaa, /* PA_SC_EDGERULE */ + 0x00000000, /* PA_SU_HARDWARE_SCREEN_OFFSET */ + 0x0000000f, /* CB_TARGET_MASK */ + 0x0000000f, /* CB_SHADER_MASK */ + + 0xc0226900, + 0x00000094, + 0x80000000, /* PA_SC_VPORT_SCISSOR_0_TL */ + 0x20002000, /* PA_SC_VPORT_SCISSOR_0_BR */ + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x00000000, /* PA_SC_VPORT_ZMIN_0 */ + 0x3f800000, /* PA_SC_VPORT_ZMAX_0 */ + + 0xc0026900, + 0x000000d9, + 0x00000000, /* CP_RINGID */ + 0x00000000, /* CP_VMID */ + + 0xc0046900, + 0x00000100, + 0xffffffff, /* VGT_MAX_VTX_INDX */ + 0x00000000, /* VGT_MIN_VTX_INDX */ + 0x00000000, /* VGT_INDX_OFFSET */ + 0x00000000, /* VGT_MULTI_PRIM_IB_RESET_INDX */ + + 0xc0046900, + 0x00000105, + 0x00000000, /* CB_BLEND_RED */ + 0x00000000, /* CB_BLEND_GREEN */ + 0x00000000, /* CB_BLEND_BLUE */ + 0x00000000, /* CB_BLEND_ALPHA */ + + 0xc0016900, + 0x000001e0, + 0x00000000, /* CB_BLEND0_CONTROL */ + + 0xc00e6900, + 0x00000200, + 0x00000000, /* DB_DEPTH_CONTROL */ + 0x00000000, /* DB_EQAA */ + 0x00cc0010, /* CB_COLOR_CONTROL */ + 0x00000210, /* DB_SHADER_CONTROL */ + 0x00010000, /* PA_CL_CLIP_CNTL */ + 0x00000004, /* PA_SU_SC_MODE_CNTL */ + 0x00000100, /* PA_CL_VTE_CNTL */ + 0x00000000, /* PA_CL_VS_OUT_CNTL */ + 0x00000000, /* PA_CL_NANINF_CNTL */ + 0x00000000, /* PA_SU_LINE_STIPPLE_CNTL */ + 0x00000000, /* PA_SU_LINE_STIPPLE_SCALE */ + 0x00000000, /* PA_SU_PRIM_FILTER_CNTL */ + 0x00000000, /* */ + 0x00000000, /* */ + + 0xc0116900, + 0x00000280, + 0x00000000, /* PA_SU_POINT_SIZE */ + 0x00000000, /* PA_SU_POINT_MINMAX */ + 0x00000008, /* PA_SU_LINE_CNTL */ + 0x00000000, /* PA_SC_LINE_STIPPLE */ + 0x00000000, /* VGT_OUTPUT_PATH_CNTL */ + 0x00000000, /* VGT_HOS_CNTL */ + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, /* VGT_GS_MODE */ + + 0xc0026900, + 0x00000292, + 0x00000000, /* PA_SC_MODE_CNTL_0 */ + 0x00000000, /* PA_SC_MODE_CNTL_1 */ + + 0xc0016900, + 0x000002a1, + 0x00000000, /* VGT_PRIMITIVEID_EN */ + + 0xc0016900, + 0x000002a5, + 0x00000000, /* VGT_MULTI_PRIM_IB_RESET_EN */ + + 0xc0026900, + 0x000002a8, + 0x00000000, /* VGT_INSTANCE_STEP_RATE_0 */ + 0x00000000, + + 0xc0026900, + 0x000002ad, + 0x00000000, /* VGT_REUSE_OFF */ + 0x00000000, + + 0xc0016900, + 0x000002d5, + 0x00000000, /* VGT_SHADER_STAGES_EN */ + + 0xc0016900, + 0x000002dc, + 0x0000aa00, /* DB_ALPHA_TO_MASK */ + + 0xc0066900, + 0x000002de, + 0x00000000, /* PA_SU_POLY_OFFSET_DB_FMT_CNTL */ + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + + 0xc0026900, + 0x000002e5, + 0x00000000, /* VGT_STRMOUT_CONFIG */ + 0x00000000, + + 0xc01b6900, + 0x000002f5, + 0x76543210, /* PA_SC_CENTROID_PRIORITY_0 */ + 0xfedcba98, /* PA_SC_CENTROID_PRIORITY_1 */ + 0x00000000, /* PA_SC_LINE_CNTL */ + 0x00000000, /* PA_SC_AA_CONFIG */ + 0x00000005, /* PA_SU_VTX_CNTL */ + 0x3f800000, /* PA_CL_GB_VERT_CLIP_ADJ */ + 0x3f800000, /* PA_CL_GB_VERT_DISC_ADJ */ + 0x3f800000, /* PA_CL_GB_HORZ_CLIP_ADJ */ + 0x3f800000, /* PA_CL_GB_HORZ_DISC_ADJ */ + 0x00000000, /* PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_0 */ + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0xffffffff, /* PA_SC_AA_MASK_X0Y0_X1Y0 */ + 0xffffffff, + + 0xc0026900, + 0x00000316, + 0x0000000e, /* VGT_VERTEX_REUSE_BLOCK_CNTL */ + 0x00000010, /* */ +}; + +static const u32 si_default_size = ARRAY_SIZE(si_default_state); #endif -- cgit v1.2.3 From d1826081bb31f1ed5daa1ba5bb7f0daf51afc2a7 Mon Sep 17 00:00:00 2001 From: Grigory Vasilyev Date: Tue, 5 Apr 2022 02:33:04 +0300 Subject: drm/amdgpu: Remove leftover igp_lane_info MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Variable igp_lane_info always is 0. 0 & any value = 0 and false. In this way, all сonditional statements will false. The code was leftover from when the code was ported from radeon where igp_lane_info was derived from the vbios on supported platforms. [update commit message - Alex] Signed-off-by: Grigory Vasilyev Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/atombios_encoders.c | 21 --------------------- 1 file changed, 21 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/atombios_encoders.c b/drivers/gpu/drm/amd/amdgpu/atombios_encoders.c index a92d86e12718..d4f5a584075d 100644 --- a/drivers/gpu/drm/amd/amdgpu/atombios_encoders.c +++ b/drivers/gpu/drm/amd/amdgpu/atombios_encoders.c @@ -765,7 +765,6 @@ amdgpu_atombios_encoder_setup_dig_transmitter(struct drm_encoder *encoder, int a int dp_clock = 0; int dp_lane_count = 0; int connector_object_id = 0; - int igp_lane_info = 0; int dig_encoder = dig->dig_encoder; int hpd_id = AMDGPU_HPD_NONE; @@ -848,26 +847,6 @@ amdgpu_atombios_encoder_setup_dig_transmitter(struct drm_encoder *encoder, int a else args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_DIG1_ENCODER; - if ((adev->flags & AMD_IS_APU) && - (amdgpu_encoder->encoder_id == ENCODER_OBJECT_ID_INTERNAL_UNIPHY)) { - if (is_dp || - !amdgpu_dig_monitor_is_duallink(encoder, amdgpu_encoder->pixel_clock)) { - if (igp_lane_info & 0x1) - args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_LANE_0_3; - else if (igp_lane_info & 0x2) - args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_LANE_4_7; - else if (igp_lane_info & 0x4) - args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_LANE_8_11; - else if (igp_lane_info & 0x8) - args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_LANE_12_15; - } else { - if (igp_lane_info & 0x3) - args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_LANE_0_7; - else if (igp_lane_info & 0xc) - args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_LANE_8_15; - } - } - if (dig->linkb) args.v1.ucConfig |= ATOM_TRANSMITTER_CONFIG_LINKB; else -- cgit v1.2.3 From 60f6fe665e854f94486aac6251950d62b1356a48 Mon Sep 17 00:00:00 2001 From: Dmytro Laktyushkin Date: Wed, 30 Mar 2022 16:05:50 -0400 Subject: drm/amd/display: update dcn315 clock table read [Why & How] Make dcn315 base its clock table off dcfclk rather than fclk. This change also adds some sanity checking to make sure an empty pmfw table does not result in invalid dal clocks. Reviewed-by: Charlene Liu Acked-by: Qingqing Zhuo Signed-off-by: Dmytro Laktyushkin Signed-off-by: Alex Deucher --- .../amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c | 107 +++++++++++++-------- 1 file changed, 67 insertions(+), 40 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c index edda572dc570..8be4c1970628 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c @@ -436,57 +436,84 @@ static void dcn315_clk_mgr_helper_populate_bw_params( struct integrated_info *bios_info, const DpmClocks_315_t *clock_table) { - int i, j; + int i; struct clk_bw_params *bw_params = clk_mgr->base.bw_params; - uint32_t max_dispclk = 0, max_dppclk = 0; - - j = -1; - - ASSERT(NUM_DF_PSTATE_LEVELS <= MAX_NUM_DPM_LVL); - - /* Find lowest DPM, FCLK is filled in reverse order*/ - - for (i = NUM_DF_PSTATE_LEVELS - 1; i >= 0; i--) { - if (clock_table->DfPstateTable[i].FClk != 0) { - j = i; - break; + uint32_t max_dispclk, max_dppclk, max_pstate, max_socclk, max_fclk = 0, min_pstate = 0; + struct clk_limit_table_entry def_max = bw_params->clk_table.entries[bw_params->clk_table.num_entries - 1]; + + max_dispclk = find_max_clk_value(clock_table->DispClocks, clock_table->NumDispClkLevelsEnabled); + max_dppclk = find_max_clk_value(clock_table->DppClocks, clock_table->NumDispClkLevelsEnabled); + max_socclk = find_max_clk_value(clock_table->SocClocks, clock_table->NumSocClkLevelsEnabled); + + /* Find highest fclk pstate */ + for (i = 0; i < clock_table->NumDfPstatesEnabled; i++) { + if (clock_table->DfPstateTable[i].FClk > max_fclk) { + max_fclk = clock_table->DfPstateTable[i].FClk; + max_pstate = i; } } - if (j == -1) { - /* clock table is all 0s, just use our own hardcode */ - ASSERT(0); - return; - } - - bw_params->clk_table.num_entries = j + 1; - - /* dispclk and dppclk can be max at any voltage, same number of levels for both */ - if (clock_table->NumDispClkLevelsEnabled <= NUM_DISPCLK_DPM_LEVELS && - clock_table->NumDispClkLevelsEnabled <= NUM_DPPCLK_DPM_LEVELS) { - max_dispclk = find_max_clk_value(clock_table->DispClocks, clock_table->NumDispClkLevelsEnabled); - max_dppclk = find_max_clk_value(clock_table->DppClocks, clock_table->NumDispClkLevelsEnabled); - } else { - ASSERT(0); - } + /* For 315 we want to base clock table on dcfclk, need at least one entry regardless of pmfw table */ + for (i = 0; i < clock_table->NumDcfClkLevelsEnabled; i++) { + int j; + uint32_t min_fclk = clock_table->DfPstateTable[0].FClk; - for (i = 0; i < bw_params->clk_table.num_entries; i++, j--) { - int temp; + for (j = 1; j < clock_table->NumDfPstatesEnabled; j++) { + if (clock_table->DfPstateTable[j].Voltage <= clock_table->SocVoltage[i] + && clock_table->DfPstateTable[j].FClk < min_fclk) { + min_fclk = clock_table->DfPstateTable[j].FClk; + min_pstate = j; + } + } - bw_params->clk_table.entries[i].fclk_mhz = clock_table->DfPstateTable[j].FClk; - bw_params->clk_table.entries[i].memclk_mhz = clock_table->DfPstateTable[j].MemClk; - bw_params->clk_table.entries[i].voltage = clock_table->DfPstateTable[j].Voltage; + bw_params->clk_table.entries[i].fclk_mhz = min_fclk; + bw_params->clk_table.entries[i].memclk_mhz = clock_table->DfPstateTable[min_pstate].MemClk; + bw_params->clk_table.entries[i].voltage = clock_table->DfPstateTable[min_pstate].Voltage; + bw_params->clk_table.entries[i].dcfclk_mhz = clock_table->DcfClocks[i]; + bw_params->clk_table.entries[i].socclk_mhz = clock_table->SocClocks[i]; + bw_params->clk_table.entries[i].dispclk_mhz = max_dispclk; + bw_params->clk_table.entries[i].dppclk_mhz = max_dppclk; bw_params->clk_table.entries[i].wck_ratio = 1; - temp = find_clk_for_voltage(clock_table, clock_table->DcfClocks, clock_table->DfPstateTable[j].Voltage); - if (temp) - bw_params->clk_table.entries[i].dcfclk_mhz = temp; - temp = find_clk_for_voltage(clock_table, clock_table->SocClocks, clock_table->DfPstateTable[j].Voltage); - if (temp) - bw_params->clk_table.entries[i].socclk_mhz = temp; + }; + + /* Make sure to include at least one entry and highest pstate */ + if (max_pstate != min_pstate) { + bw_params->clk_table.entries[i].fclk_mhz = max_fclk; + bw_params->clk_table.entries[i].memclk_mhz = clock_table->DfPstateTable[max_pstate].MemClk; + bw_params->clk_table.entries[i].voltage = clock_table->DfPstateTable[max_pstate].Voltage; + bw_params->clk_table.entries[i].dcfclk_mhz = find_clk_for_voltage( + clock_table, clock_table->DcfClocks, clock_table->DfPstateTable[max_pstate].Voltage); + bw_params->clk_table.entries[i].socclk_mhz = find_clk_for_voltage( + clock_table, clock_table->SocClocks, clock_table->DfPstateTable[max_pstate].Voltage); bw_params->clk_table.entries[i].dispclk_mhz = max_dispclk; bw_params->clk_table.entries[i].dppclk_mhz = max_dppclk; + bw_params->clk_table.entries[i].wck_ratio = 1; + i++; } + bw_params->clk_table.num_entries = i; + + /* Include highest socclk */ + if (bw_params->clk_table.entries[i-1].socclk_mhz < max_socclk) + bw_params->clk_table.entries[i-1].socclk_mhz = max_socclk; + /* Set any 0 clocks to max default setting. Not an issue for + * power since we aren't doing switching in such case anyway + */ + for (i = 0; i < bw_params->clk_table.num_entries; i++) { + if (!bw_params->clk_table.entries[i].fclk_mhz) { + bw_params->clk_table.entries[i].fclk_mhz = def_max.fclk_mhz; + bw_params->clk_table.entries[i].memclk_mhz = def_max.memclk_mhz; + bw_params->clk_table.entries[i].voltage = def_max.voltage; + } + if (!bw_params->clk_table.entries[i].dcfclk_mhz) + bw_params->clk_table.entries[i].dcfclk_mhz = def_max.dcfclk_mhz; + if (!bw_params->clk_table.entries[i].socclk_mhz) + bw_params->clk_table.entries[i].socclk_mhz = def_max.socclk_mhz; + if (!bw_params->clk_table.entries[i].dispclk_mhz) + bw_params->clk_table.entries[i].dispclk_mhz = def_max.dispclk_mhz; + if (!bw_params->clk_table.entries[i].dppclk_mhz) + bw_params->clk_table.entries[i].dppclk_mhz = def_max.dppclk_mhz; + } bw_params->vram_type = bios_info->memory_type; bw_params->num_channels = bios_info->ma_channel_number; -- cgit v1.2.3 From 9bbcf182060842a652c30f63692c6b6a8eb54386 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Fri, 1 Apr 2022 11:08:48 -0400 Subject: drm/amdgpu/smu10: fix SoC/fclk units in auto mode SMU takes clock limits in Mhz units. socclk and fclk were using 10 khz units in some cases. Switch to Mhz units. Fixes higher than required SoC clocks. Fixes: 97cf32996c46d9 ("drm/amd/pm: Removed fixed clock in auto mode DPM") Reviewed-by: Paul Menzel Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu10_hwmgr.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu10_hwmgr.c index 9ddd8491ff00..ede71de2343d 100644 --- a/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu10_hwmgr.c +++ b/drivers/gpu/drm/amd/pm/powerplay/hwmgr/smu10_hwmgr.c @@ -773,13 +773,13 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr, smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetHardMinFclkByFreq, hwmgr->display_config->num_display > 3 ? - data->clock_vol_info.vdd_dep_on_fclk->entries[0].clk : + (data->clock_vol_info.vdd_dep_on_fclk->entries[0].clk / 100) : min_mclk, NULL); smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetHardMinSocclkByFreq, - data->clock_vol_info.vdd_dep_on_socclk->entries[0].clk, + data->clock_vol_info.vdd_dep_on_socclk->entries[0].clk / 100, NULL); smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetHardMinVcn, @@ -792,11 +792,11 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr, NULL); smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetSoftMaxFclkByFreq, - data->clock_vol_info.vdd_dep_on_fclk->entries[index_fclk].clk, + data->clock_vol_info.vdd_dep_on_fclk->entries[index_fclk].clk / 100, NULL); smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetSoftMaxSocclkByFreq, - data->clock_vol_info.vdd_dep_on_socclk->entries[index_socclk].clk, + data->clock_vol_info.vdd_dep_on_socclk->entries[index_socclk].clk / 100, NULL); smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetSoftMaxVcn, -- cgit v1.2.3 From 8d2aad983de2a332bf8c22798ab6799f06864fed Mon Sep 17 00:00:00 2001 From: Lv Ruyi Date: Wed, 6 Apr 2022 07:28:40 +0000 Subject: drm/amd/dc: remove duplicate include 'dm_services.h' included in 'freesync,c' is duplicated, so remove one. Reported-by: Zeal Robot Signed-off-by: Lv Ruyi Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/modules/freesync/freesync.c | 1 - 1 file changed, 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c index d72566c6928a..0130f1879116 100644 --- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c +++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c @@ -29,7 +29,6 @@ #include "dc.h" #include "mod_freesync.h" #include "core_types.h" -#include "dm_services.h" #define MOD_FREESYNC_MAX_CONCURRENT_STREAMS 32 -- cgit v1.2.3 From 5273e82c5f47fff94058ff8ee002650476e24719 Mon Sep 17 00:00:00 2001 From: Felix Kuehling Date: Tue, 1 Mar 2022 20:40:45 -0500 Subject: drm/amdkfd: Improve concurrency of event handling Use rcu_read_lock to read p->event_idr concurrently with other readers and writers. Use p->event_mutex only for creating and destroying events and in kfd_wait_on_events. Protect the contents of the kfd_event structure with a per-event spinlock that can be taken inside the rcu_read_lock critical section. This eliminates contention of p->event_mutex in set_event, which tends to be on the critical path for dispatch latency even when busy waiting is used. It also eliminates lock contention in event interrupt handlers. Since the p->event_mutex is now used much less, the impact of requiring it in kfd_wait_on_events should also be much smaller. This should improve event handling latency for processes using multiple GPUs concurrently. v2: Reschedule the worker periodically to avoid soft lockup warnings Signed-off-by: Felix Kuehling Reviewed-by: Sean Keely # v1 Tested-by: Sanjay Tripathi Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_events.c | 119 +++++++++++++++++++---------- drivers/gpu/drm/amd/amdkfd/kfd_events.h | 1 + drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c | 11 ++- 3 files changed, 88 insertions(+), 43 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c index 64f4a51cc880..0fef24b0b915 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c @@ -128,8 +128,8 @@ static int allocate_event_notification_slot(struct kfd_process *p, } /* - * Assumes that p->event_mutex is held and of course that p is not going - * away (current or locked). + * Assumes that p->event_mutex or rcu_readlock is held and of course that p is + * not going away. */ static struct kfd_event *lookup_event_by_id(struct kfd_process *p, uint32_t id) { @@ -251,15 +251,18 @@ static void destroy_event(struct kfd_process *p, struct kfd_event *ev) struct kfd_event_waiter *waiter; /* Wake up pending waiters. They will return failure */ + spin_lock(&ev->lock); list_for_each_entry(waiter, &ev->wq.head, wait.entry) - waiter->event = NULL; + WRITE_ONCE(waiter->event, NULL); wake_up_all(&ev->wq); + spin_unlock(&ev->lock); if (ev->type == KFD_EVENT_TYPE_SIGNAL || ev->type == KFD_EVENT_TYPE_DEBUG) p->signal_event_count--; idr_remove(&p->event_idr, ev->event_id); + synchronize_rcu(); kfree(ev); } @@ -392,6 +395,7 @@ int kfd_event_create(struct file *devkfd, struct kfd_process *p, ev->auto_reset = auto_reset; ev->signaled = false; + spin_lock_init(&ev->lock); init_waitqueue_head(&ev->wq); *event_page_offset = 0; @@ -466,6 +470,7 @@ int kfd_criu_restore_event(struct file *devkfd, ev->auto_reset = ev_priv->auto_reset; ev->signaled = ev_priv->signaled; + spin_lock_init(&ev->lock); init_waitqueue_head(&ev->wq); mutex_lock(&p->event_mutex); @@ -609,13 +614,13 @@ static void set_event(struct kfd_event *ev) /* Auto reset if the list is non-empty and we're waking * someone. waitqueue_active is safe here because we're - * protected by the p->event_mutex, which is also held when + * protected by the ev->lock, which is also held when * updating the wait queues in kfd_wait_on_events. */ ev->signaled = !ev->auto_reset || !waitqueue_active(&ev->wq); list_for_each_entry(waiter, &ev->wq.head, wait.entry) - waiter->activated = true; + WRITE_ONCE(waiter->activated, true); wake_up_all(&ev->wq); } @@ -626,16 +631,18 @@ int kfd_set_event(struct kfd_process *p, uint32_t event_id) int ret = 0; struct kfd_event *ev; - mutex_lock(&p->event_mutex); + rcu_read_lock(); ev = lookup_event_by_id(p, event_id); + spin_lock(&ev->lock); if (ev && event_can_be_cpu_signaled(ev)) set_event(ev); else ret = -EINVAL; - mutex_unlock(&p->event_mutex); + spin_unlock(&ev->lock); + rcu_read_unlock(); return ret; } @@ -650,23 +657,25 @@ int kfd_reset_event(struct kfd_process *p, uint32_t event_id) int ret = 0; struct kfd_event *ev; - mutex_lock(&p->event_mutex); + rcu_read_lock(); ev = lookup_event_by_id(p, event_id); + spin_lock(&ev->lock); if (ev && event_can_be_cpu_signaled(ev)) reset_event(ev); else ret = -EINVAL; - mutex_unlock(&p->event_mutex); + spin_unlock(&ev->lock); + rcu_read_unlock(); return ret; } static void acknowledge_signal(struct kfd_process *p, struct kfd_event *ev) { - page_slots(p->signal_page)[ev->event_id] = UNSIGNALED_EVENT_SLOT; + WRITE_ONCE(page_slots(p->signal_page)[ev->event_id], UNSIGNALED_EVENT_SLOT); } static void set_event_from_interrupt(struct kfd_process *p, @@ -674,7 +683,9 @@ static void set_event_from_interrupt(struct kfd_process *p, { if (ev && event_can_be_gpu_signaled(ev)) { acknowledge_signal(p, ev); + spin_lock(&ev->lock); set_event(ev); + spin_unlock(&ev->lock); } } @@ -693,7 +704,7 @@ void kfd_signal_event_interrupt(u32 pasid, uint32_t partial_id, if (!p) return; /* Presumably process exited. */ - mutex_lock(&p->event_mutex); + rcu_read_lock(); if (valid_id_bits) ev = lookup_signaled_event_by_partial_id(p, partial_id, @@ -721,7 +732,7 @@ void kfd_signal_event_interrupt(u32 pasid, uint32_t partial_id, if (id >= KFD_SIGNAL_EVENT_LIMIT) break; - if (slots[id] != UNSIGNALED_EVENT_SLOT) + if (READ_ONCE(slots[id]) != UNSIGNALED_EVENT_SLOT) set_event_from_interrupt(p, ev); } } else { @@ -730,14 +741,14 @@ void kfd_signal_event_interrupt(u32 pasid, uint32_t partial_id, * only signaled events from the IDR. */ for (id = 0; id < KFD_SIGNAL_EVENT_LIMIT; id++) - if (slots[id] != UNSIGNALED_EVENT_SLOT) { + if (READ_ONCE(slots[id]) != UNSIGNALED_EVENT_SLOT) { ev = lookup_event_by_id(p, id); set_event_from_interrupt(p, ev); } } } - mutex_unlock(&p->event_mutex); + rcu_read_unlock(); kfd_unref_process(p); } @@ -769,9 +780,11 @@ static int init_event_waiter_get_status(struct kfd_process *p, if (!ev) return -EINVAL; + spin_lock(&ev->lock); waiter->event = ev; waiter->activated = ev->signaled; ev->signaled = ev->signaled && !ev->auto_reset; + spin_unlock(&ev->lock); return 0; } @@ -783,8 +796,11 @@ static void init_event_waiter_add_to_waitlist(struct kfd_event_waiter *waiter) /* Only add to the wait list if we actually need to * wait on this event. */ - if (!waiter->activated) + if (!waiter->activated) { + spin_lock(&ev->lock); add_wait_queue(&ev->wq, &waiter->wait); + spin_unlock(&ev->lock); + } } /* test_event_condition - Test condition of events being waited for @@ -804,10 +820,10 @@ static uint32_t test_event_condition(bool all, uint32_t num_events, uint32_t activated_count = 0; for (i = 0; i < num_events; i++) { - if (!event_waiters[i].event) + if (!READ_ONCE(event_waiters[i].event)) return KFD_IOC_WAIT_RESULT_FAIL; - if (event_waiters[i].activated) { + if (READ_ONCE(event_waiters[i].activated)) { if (!all) return KFD_IOC_WAIT_RESULT_COMPLETE; @@ -836,6 +852,8 @@ static int copy_signaled_event_data(uint32_t num_events, for (i = 0; i < num_events; i++) { waiter = &event_waiters[i]; event = waiter->event; + if (!event) + return -EINVAL; /* event was destroyed */ if (waiter->activated && event->type == KFD_EVENT_TYPE_MEMORY) { dst = &data[i].memory_exception_data; src = &event->memory_exception_data; @@ -846,11 +864,8 @@ static int copy_signaled_event_data(uint32_t num_events, } return 0; - } - - static long user_timeout_to_jiffies(uint32_t user_timeout_ms) { if (user_timeout_ms == KFD_EVENT_TIMEOUT_IMMEDIATE) @@ -874,9 +889,12 @@ static void free_waiters(uint32_t num_events, struct kfd_event_waiter *waiters) uint32_t i; for (i = 0; i < num_events; i++) - if (waiters[i].event) + if (waiters[i].event) { + spin_lock(&waiters[i].event->lock); remove_wait_queue(&waiters[i].event->wq, &waiters[i].wait); + spin_unlock(&waiters[i].event->lock); + } kfree(waiters); } @@ -900,6 +918,9 @@ int kfd_wait_on_events(struct kfd_process *p, goto out; } + /* Use p->event_mutex here to protect against concurrent creation and + * destruction of events while we initialize event_waiters. + */ mutex_lock(&p->event_mutex); for (i = 0; i < num_events; i++) { @@ -978,14 +999,19 @@ int kfd_wait_on_events(struct kfd_process *p, } __set_current_state(TASK_RUNNING); + mutex_lock(&p->event_mutex); /* copy_signaled_event_data may sleep. So this has to happen * after the task state is set back to RUNNING. + * + * The event may also have been destroyed after signaling. So + * copy_signaled_event_data also must confirm that the event + * still exists. Therefore this must be under the p->event_mutex + * which is also held when events are destroyed. */ if (!ret && *wait_result == KFD_IOC_WAIT_RESULT_COMPLETE) ret = copy_signaled_event_data(num_events, event_waiters, events); - mutex_lock(&p->event_mutex); out_unlock: free_waiters(num_events, event_waiters); mutex_unlock(&p->event_mutex); @@ -1044,8 +1070,7 @@ int kfd_event_mmap(struct kfd_process *p, struct vm_area_struct *vma) } /* - * Assumes that p->event_mutex is held and of course - * that p is not going away (current or locked). + * Assumes that p is not going away. */ static void lookup_events_by_type_and_signal(struct kfd_process *p, int type, void *event_data) @@ -1057,6 +1082,8 @@ static void lookup_events_by_type_and_signal(struct kfd_process *p, ev_data = (struct kfd_hsa_memory_exception_data *) event_data; + rcu_read_lock(); + id = KFD_FIRST_NONSIGNAL_EVENT_ID; idr_for_each_entry_continue(&p->event_idr, ev, id) if (ev->type == type) { @@ -1064,9 +1091,11 @@ static void lookup_events_by_type_and_signal(struct kfd_process *p, dev_dbg(kfd_device, "Event found: id %X type %d", ev->event_id, ev->type); + spin_lock(&ev->lock); set_event(ev); if (ev->type == KFD_EVENT_TYPE_MEMORY && ev_data) ev->memory_exception_data = *ev_data; + spin_unlock(&ev->lock); } if (type == KFD_EVENT_TYPE_MEMORY) { @@ -1089,6 +1118,8 @@ static void lookup_events_by_type_and_signal(struct kfd_process *p, p->lead_thread->pid, p->pasid); } } + + rcu_read_unlock(); } #ifdef KFD_SUPPORT_IOMMU_V2 @@ -1164,16 +1195,10 @@ void kfd_signal_iommu_event(struct kfd_dev *dev, u32 pasid, if (KFD_GC_VERSION(dev) != IP_VERSION(9, 1, 0) && KFD_GC_VERSION(dev) != IP_VERSION(9, 2, 2) && - KFD_GC_VERSION(dev) != IP_VERSION(9, 3, 0)) { - mutex_lock(&p->event_mutex); - - /* Lookup events by type and signal them */ + KFD_GC_VERSION(dev) != IP_VERSION(9, 3, 0)) lookup_events_by_type_and_signal(p, KFD_EVENT_TYPE_MEMORY, &memory_exception_data); - mutex_unlock(&p->event_mutex); - } - kfd_unref_process(p); } #endif /* KFD_SUPPORT_IOMMU_V2 */ @@ -1190,12 +1215,7 @@ void kfd_signal_hw_exception_event(u32 pasid) if (!p) return; /* Presumably process exited. */ - mutex_lock(&p->event_mutex); - - /* Lookup events by type and signal them */ lookup_events_by_type_and_signal(p, KFD_EVENT_TYPE_HW_EXCEPTION, NULL); - - mutex_unlock(&p->event_mutex); kfd_unref_process(p); } @@ -1231,16 +1251,19 @@ void kfd_signal_vm_fault_event(struct kfd_dev *dev, u32 pasid, info->prot_write ? 1 : 0; memory_exception_data.failure.imprecise = 0; } - mutex_lock(&p->event_mutex); + + rcu_read_lock(); id = KFD_FIRST_NONSIGNAL_EVENT_ID; idr_for_each_entry_continue(&p->event_idr, ev, id) if (ev->type == KFD_EVENT_TYPE_MEMORY) { + spin_lock(&ev->lock); ev->memory_exception_data = memory_exception_data; set_event(ev); + spin_unlock(&ev->lock); } - mutex_unlock(&p->event_mutex); + rcu_read_unlock(); kfd_unref_process(p); } @@ -1274,22 +1297,28 @@ void kfd_signal_reset_event(struct kfd_dev *dev) continue; } - mutex_lock(&p->event_mutex); + rcu_read_lock(); + id = KFD_FIRST_NONSIGNAL_EVENT_ID; idr_for_each_entry_continue(&p->event_idr, ev, id) { if (ev->type == KFD_EVENT_TYPE_HW_EXCEPTION) { + spin_lock(&ev->lock); ev->hw_exception_data = hw_exception_data; ev->hw_exception_data.gpu_id = user_gpu_id; set_event(ev); + spin_unlock(&ev->lock); } if (ev->type == KFD_EVENT_TYPE_MEMORY && reset_cause == KFD_HW_EXCEPTION_ECC) { + spin_lock(&ev->lock); ev->memory_exception_data = memory_exception_data; ev->memory_exception_data.gpu_id = user_gpu_id; set_event(ev); + spin_unlock(&ev->lock); } } - mutex_unlock(&p->event_mutex); + + rcu_read_unlock(); } srcu_read_unlock(&kfd_processes_srcu, idx); } @@ -1322,19 +1351,25 @@ void kfd_signal_poison_consumed_event(struct kfd_dev *dev, u32 pasid) memory_exception_data.gpu_id = user_gpu_id; memory_exception_data.failure.imprecise = true; - mutex_lock(&p->event_mutex); + rcu_read_lock(); + idr_for_each_entry_continue(&p->event_idr, ev, id) { if (ev->type == KFD_EVENT_TYPE_HW_EXCEPTION) { + spin_lock(&ev->lock); ev->hw_exception_data = hw_exception_data; set_event(ev); + spin_unlock(&ev->lock); } if (ev->type == KFD_EVENT_TYPE_MEMORY) { + spin_lock(&ev->lock); ev->memory_exception_data = memory_exception_data; set_event(ev); + spin_unlock(&ev->lock); } } - mutex_unlock(&p->event_mutex); + + rcu_read_unlock(); /* user application will handle SIGBUS signal */ send_sig(SIGBUS, p->lead_thread, 0); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.h b/drivers/gpu/drm/amd/amdkfd/kfd_events.h index 1238af11916e..55d376f56021 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.h @@ -59,6 +59,7 @@ struct kfd_event { int type; + spinlock_t lock; wait_queue_head_t wq; /* List of event waiters. */ /* Only for signal events. */ diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c index 9178cfe34f20..a9466d154395 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c @@ -146,15 +146,24 @@ static void interrupt_wq(struct work_struct *work) struct kfd_dev *dev = container_of(work, struct kfd_dev, interrupt_work); uint32_t ih_ring_entry[KFD_MAX_RING_ENTRY_SIZE]; + long start_jiffies = jiffies; if (dev->device_info.ih_ring_entry_size > sizeof(ih_ring_entry)) { dev_err_once(dev->adev->dev, "Ring entry too small\n"); return; } - while (dequeue_ih_ring_entry(dev, ih_ring_entry)) + while (dequeue_ih_ring_entry(dev, ih_ring_entry)) { dev->device_info.event_interrupt_class->interrupt_wq(dev, ih_ring_entry); + if (jiffies - start_jiffies > HZ) { + /* If we spent more than a second processing signals, + * reschedule the worker to avoid soft-lockup warnings + */ + queue_work(dev->ih_wq, &dev->interrupt_work); + break; + } + } } bool interrupt_is_wanted(struct kfd_dev *dev, -- cgit v1.2.3 From 67229b27ed6fdb6bfce04f61ce4026c57b32ba15 Mon Sep 17 00:00:00 2001 From: Melissa Wen Date: Tue, 29 Mar 2022 19:18:35 -0100 Subject: drm/amd/display: don't ignore alpha property on pre-multiplied mode "Pre-multiplied" is the default pixel blend mode for KMS/DRM, as documented in supported_modes of drm_plane_create_blend_mode_property(): https://cgit.freedesktop.org/drm/drm-misc/tree/drivers/gpu/drm/drm_blend.c In this mode, both 'pixel alpha' and 'plane alpha' participate in the calculation, as described by the pixel blend mode formula in KMS/DRM documentation: out.rgb = plane_alpha * fg.rgb + (1 - (plane_alpha * fg.alpha)) * bg.rgb Considering the blend config mechanisms we have in the driver so far, the alpha mode that better fits this blend mode is the _PER_PIXEL_ALPHA_COMBINED_GLOBAL_GAIN, where the value for global_gain is the plane alpha (global_alpha). With this change, alpha property stops to be ignored. It also addresses Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1734 v2: * keep the 8-bit value for global_alpha_value (Nicholas) * correct the logical ordering for combined global gain (Nicholas) * apply to dcn10 too (Nicholas) Signed-off-by: Melissa Wen Tested-by: Rodrigo Siqueira Reviewed-by: Harry Wentland Tested-by: Simon Ser Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 14 +++++++++----- drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c | 14 +++++++++----- 2 files changed, 18 insertions(+), 10 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index 50820e79d3c4..724c04fe004b 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c @@ -2526,14 +2526,18 @@ void dcn10_update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx) struct mpc *mpc = dc->res_pool->mpc; struct mpc_tree *mpc_tree_params = &(pipe_ctx->stream_res.opp->mpc_tree_params); - if (per_pixel_alpha) - blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_PER_PIXEL_ALPHA; - else - blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_GLOBAL_ALPHA; - blnd_cfg.overlap_only = false; blnd_cfg.global_gain = 0xff; + if (per_pixel_alpha && pipe_ctx->plane_state->global_alpha) { + blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_PER_PIXEL_ALPHA_COMBINED_GLOBAL_GAIN; + blnd_cfg.global_gain = pipe_ctx->plane_state->global_alpha_value; + } else if (per_pixel_alpha) { + blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_PER_PIXEL_ALPHA; + } else { + blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_GLOBAL_ALPHA; + } + if (pipe_ctx->plane_state->global_alpha) blnd_cfg.global_alpha = pipe_ctx->plane_state->global_alpha_value; else diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c index 4290eaf11a04..b627c41713cc 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c @@ -2344,14 +2344,18 @@ void dcn20_update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx) struct mpc *mpc = dc->res_pool->mpc; struct mpc_tree *mpc_tree_params = &(pipe_ctx->stream_res.opp->mpc_tree_params); - if (per_pixel_alpha) - blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_PER_PIXEL_ALPHA; - else - blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_GLOBAL_ALPHA; - blnd_cfg.overlap_only = false; blnd_cfg.global_gain = 0xff; + if (per_pixel_alpha && pipe_ctx->plane_state->global_alpha) { + blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_PER_PIXEL_ALPHA_COMBINED_GLOBAL_GAIN; + blnd_cfg.global_gain = pipe_ctx->plane_state->global_alpha_value; + } else if (per_pixel_alpha) { + blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_PER_PIXEL_ALPHA; + } else { + blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_GLOBAL_ALPHA; + } + if (pipe_ctx->plane_state->global_alpha) blnd_cfg.global_alpha = pipe_ctx->plane_state->global_alpha_value; else -- cgit v1.2.3 From dd481828975f1a7e535e51f0e10178ae8e05fe33 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 6 Apr 2022 14:39:03 -0400 Subject: drm/amdgpu: fix VCN 3.1.2 firmware name Drop the trailing vcn. Fixes: afc2f276057ea1 ("drm/amdgpu/vcn: add vcn support for vcn 3.1.2") Reviewed-by: James Zhu Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c index f99093f2ebc7..a0ee828a4a97 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c @@ -52,7 +52,7 @@ #define FIRMWARE_ALDEBARAN "amdgpu/aldebaran_vcn.bin" #define FIRMWARE_BEIGE_GOBY "amdgpu/beige_goby_vcn.bin" #define FIRMWARE_YELLOW_CARP "amdgpu/yellow_carp_vcn.bin" -#define FIRMWARE_VCN_3_1_2 "amdgpu/vcn_3_1_2_vcn.bin" +#define FIRMWARE_VCN_3_1_2 "amdgpu/vcn_3_1_2.bin" MODULE_FIRMWARE(FIRMWARE_RAVEN); MODULE_FIRMWARE(FIRMWARE_PICASSO); -- cgit v1.2.3 From 9e051720f9d37ef6f878a6106478c82fb9b375d1 Mon Sep 17 00:00:00 2001 From: Kai-Heng Feng Date: Thu, 7 Apr 2022 20:12:28 +0800 Subject: drm/amdgpu: Ensure HDA function is suspended before ASIC reset DP/HDMI audio on AMD PRO VII stops working after S3: [ 149.450391] amdgpu 0000:63:00.0: amdgpu: MODE1 reset [ 149.450395] amdgpu 0000:63:00.0: amdgpu: GPU mode1 reset [ 149.450494] amdgpu 0000:63:00.0: amdgpu: GPU psp mode1 reset [ 149.983693] snd_hda_intel 0000:63:00.1: refused to change power state from D0 to D3hot [ 150.003439] amdgpu 0000:63:00.0: refused to change power state from D0 to D3hot ... [ 155.432975] snd_hda_intel 0000:63:00.1: CORB reset timeout#2, CORBRP = 65535 The offending commit is daf8de0874ab5b ("drm/amdgpu: always reset the asic in suspend (v2)"). Commit 34452ac3038a7 ("drm/amdgpu: don't use BACO for reset in S3 ") doesn't help, so the issue is something different. Assuming that to make HDA resume to D0 fully realized, it needs to be successfully put to D3 first. And this guesswork proves working, by moving amdgpu_asic_reset() to noirq callback, so it's called after HDA function is in D3. Fixes: daf8de0874ab5b ("drm/amdgpu: always reset the asic in suspend (v2)") Signed-off-by: Kai-Heng Feng Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index b03663f42cc9..29e9419a914b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -2323,18 +2323,23 @@ 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); - int r; if (amdgpu_acpi_is_s0ix_active(adev)) adev->in_s0ix = true; else adev->in_s3 = true; - r = amdgpu_device_suspend(drm_dev, true); - if (r) - return r; + return amdgpu_device_suspend(drm_dev, true); +} + +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); + if (!adev->in_s0ix) - r = amdgpu_asic_reset(adev); - return r; + return amdgpu_asic_reset(adev); + + return 0; } static int amdgpu_pmops_resume(struct device *dev) @@ -2575,6 +2580,7 @@ static const struct dev_pm_ops amdgpu_pm_ops = { .prepare = amdgpu_pmops_prepare, .complete = amdgpu_pmops_complete, .suspend = amdgpu_pmops_suspend, + .suspend_noirq = amdgpu_pmops_suspend_noirq, .resume = amdgpu_pmops_resume, .freeze = amdgpu_pmops_freeze, .thaw = amdgpu_pmops_thaw, -- cgit v1.2.3 From 9b6a1ec7924e9feee70ece3f779bc058ff25f716 Mon Sep 17 00:00:00 2001 From: Tomasz Moń Date: Wed, 6 Apr 2022 21:49:21 +0200 Subject: drm/amdgpu: Enable gfxoff quirk on MacBook Pro MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Enabling gfxoff quirk results in perfectly usable graphical user interface on MacBook Pro (15-inch, 2019) with Radeon Pro Vega 20 4 GB. Without the quirk, X server is completely unusable as every few seconds there is gpu reset due to ring gfx timeout. Signed-off-by: Tomasz Moń Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 46d4bf27ebbb..b8cfcc6b1125 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -1205,6 +1205,8 @@ static const struct amdgpu_gfxoff_quirk amdgpu_gfxoff_quirk_list[] = { { 0x1002, 0x15dd, 0x103c, 0x83e7, 0xd3 }, /* GFXOFF is unstable on C6 parts with a VBIOS 113-RAVEN-114 */ { 0x1002, 0x15dd, 0x1002, 0x15dd, 0xc6 }, + /* Apple MacBook Pro (15-inch, 2019) Radeon Pro Vega 20 4 GB */ + { 0x1002, 0x69af, 0x106b, 0x019a, 0xc0 }, { 0, 0, 0, 0, 0 }, }; -- cgit v1.2.3 From dac1655cb2a92e6ba13f37dee1a56cb8a7d7f753 Mon Sep 17 00:00:00 2001 From: Tom Rix Date: Sun, 3 Apr 2022 10:21:37 -0400 Subject: drm/amd/display: cleanup extern usage in function definition Smatch reports this issue hdcp1_execution.c:500:29: warning: function 'mod_hdcp_hdcp1_dp_execution' with external linkage has definition The storage-class-specifier extern is not needed in a definition, so remove it. Reviewed-by: Harry Wentland Signed-off-by: Tom Rix Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_execution.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_execution.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_execution.c index 6ec918af3bff..1ddb4f5eac8e 100644 --- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_execution.c +++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_execution.c @@ -497,9 +497,9 @@ enum mod_hdcp_status mod_hdcp_hdcp1_execution(struct mod_hdcp *hdcp, return status; } -extern enum mod_hdcp_status mod_hdcp_hdcp1_dp_execution(struct mod_hdcp *hdcp, - struct mod_hdcp_event_context *event_ctx, - struct mod_hdcp_transition_input_hdcp1 *input) +enum mod_hdcp_status mod_hdcp_hdcp1_dp_execution(struct mod_hdcp *hdcp, + struct mod_hdcp_event_context *event_ctx, + struct mod_hdcp_transition_input_hdcp1 *input) { enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS; -- cgit v1.2.3 From 3cd3e731f34ff2f021165aeefd640acba9dd0993 Mon Sep 17 00:00:00 2001 From: Felix Kuehling Date: Wed, 6 Apr 2022 16:58:24 -0400 Subject: drm/amdkfd: Fix NULL pointer dereference Check that adev->gfx.ras is valid before using it. Fixes: 6475ae2b742876 ("drm/amdgpu: add UTCL2 RAS poison query for Aldebaran (v2)") CC: Tao Zhou Signed-off-by: Felix Kuehling Reviewed-by: Mukul Joshi Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index c18c4be1e4ac..64c6664b34e8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -727,7 +727,7 @@ void amdgpu_amdkfd_ras_poison_consumption_handler(struct amdgpu_device *adev, bo bool amdgpu_amdkfd_ras_query_utcl2_poison_status(struct amdgpu_device *adev) { - if (adev->gfx.ras->query_utcl2_poison_status) + if (adev->gfx.ras && adev->gfx.ras->query_utcl2_poison_status) return adev->gfx.ras->query_utcl2_poison_status(adev); else return false; -- cgit v1.2.3 From 49aa98ca30cd186ab33fc5802066e2024d3bfa39 Mon Sep 17 00:00:00 2001 From: Yongqiang Sun Date: Wed, 30 Mar 2022 10:48:19 -0400 Subject: drm/amd/amdgpu: Only reserve vram for firmware with vega9 MS_HYPERV host. driver loading failed on VEGA10 SRIOV VF with linux host due to a wide range of stolen reserved vram. Since VEGA10 SRIOV VF need to reserve vram for firmware with windows Hyper_V host specifically, check hypervisor type to only reserve memory for it, and the range of the reserved vram can be limited to between 5M-7M area. Fixes: faad5ccac1eaae ("drm/amdgpu: Add stolen reserved memory for MI25 SRIOV.") Signed-off-by: Yongqiang Sun Acked-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c index a66a0881a934..3e9582c245bb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c @@ -25,6 +25,7 @@ */ #include +#include #include "amdgpu.h" #include "amdgpu_gmc.h" @@ -647,11 +648,11 @@ void amdgpu_gmc_get_vbios_allocations(struct amdgpu_device *adev) case CHIP_VEGA10: adev->mman.keep_stolen_vga_memory = true; /* - * VEGA10 SRIOV VF needs some firmware reserved area. + * VEGA10 SRIOV VF with MS_HYPERV host needs some firmware reserved area. */ - if (amdgpu_sriov_vf(adev)) { - adev->mman.stolen_reserved_offset = 0x100000; - adev->mman.stolen_reserved_size = 0x600000; + if (amdgpu_sriov_vf(adev) && hypervisor_is_type(X86_HYPER_MS_HYPERV)) { + adev->mman.stolen_reserved_offset = 0x500000; + adev->mman.stolen_reserved_size = 0x200000; } break; case CHIP_RAVEN: -- cgit v1.2.3 From 25faeddcf3c3f31f9f43de5c55f7cbdd13d3ebc7 Mon Sep 17 00:00:00 2001 From: Evan Quan Date: Fri, 25 Mar 2022 18:00:02 +0800 Subject: drm/amdgpu: expand cg_flags from u32 to u64 With this, we can support more CG flags. Signed-off-by: Evan Quan Acked-by: Alex Deucher Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 6 +-- drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 5 ++- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_df.h | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 8 ++-- drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.h | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_smuio.h | 2 +- drivers/gpu/drm/amd/amdgpu/athub_v1_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/athub_v1_0.h | 2 +- drivers/gpu/drm/amd/amdgpu/athub_v2_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/athub_v2_0.h | 2 +- drivers/gpu/drm/amd/amdgpu/athub_v2_1.c | 2 +- drivers/gpu/drm/amd/amdgpu/athub_v2_1.h | 2 +- drivers/gpu/drm/amd/amdgpu/df_v1_7.c | 2 +- drivers/gpu/drm/amd/amdgpu/df_v3_6.c | 2 +- drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/hdp_v5_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c | 2 +- drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/mmhub_v2_3.c | 2 +- drivers/gpu/drm/amd/amdgpu/mmhub_v9_4.c | 2 +- drivers/gpu/drm/amd/amdgpu/navi10_ih.c | 2 +- drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c | 2 +- drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c | 2 +- drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c | 2 +- drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c | 2 +- drivers/gpu/drm/amd/amdgpu/nv.c | 2 +- drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c | 2 +- drivers/gpu/drm/amd/amdgpu/smuio_v11_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/smuio_v11_0_6.c | 2 +- drivers/gpu/drm/amd/amdgpu/smuio_v13_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/smuio_v9_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/soc15.c | 2 +- drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/vce_v3_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/vi.c | 2 +- drivers/gpu/drm/amd/include/amd_shared.h | 66 ++++++++++++++--------------- drivers/gpu/drm/amd/pm/amdgpu_pm.c | 6 +-- drivers/gpu/drm/amd/pm/inc/amdgpu_pm.h | 2 +- 53 files changed, 95 insertions(+), 92 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index cdf0818088b3..bffd24845765 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -179,7 +179,7 @@ extern int amdgpu_sched_jobs; extern int amdgpu_sched_hw_submission; extern uint amdgpu_pcie_gen_cap; extern uint amdgpu_pcie_lane_cap; -extern uint amdgpu_cg_mask; +extern u64 amdgpu_cg_mask; extern uint amdgpu_pg_mask; extern uint amdgpu_sdma_phase_quantum; extern char *amdgpu_disable_cu; @@ -322,7 +322,7 @@ int amdgpu_device_ip_set_powergating_state(void *dev, enum amd_ip_block_type block_type, enum amd_powergating_state state); void amdgpu_device_ip_get_clockgating_state(struct amdgpu_device *adev, - u32 *flags); + u64 *flags); int amdgpu_device_ip_wait_for_idle(struct amdgpu_device *adev, enum amd_ip_block_type block_type); bool amdgpu_device_ip_is_idle(struct amdgpu_device *adev, @@ -860,7 +860,7 @@ struct amdgpu_device { /* powerplay */ struct amd_powerplay powerplay; struct amdgpu_pm pm; - u32 cg_flags; + u64 cg_flags; u32 pg_flags; /* nbio */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c index 5d04d24a0d5f..1db48dc58ae9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c @@ -758,7 +758,7 @@ static ssize_t amdgpu_debugfs_gca_config_read(struct file *f, char __user *buf, /* rev==1 */ config[no_regs++] = adev->rev_id; config[no_regs++] = adev->pg_flags; - config[no_regs++] = adev->cg_flags; + config[no_regs++] = lower_32_bits(adev->cg_flags); /* rev==2 */ config[no_regs++] = adev->family; @@ -773,6 +773,9 @@ static ssize_t amdgpu_debugfs_gca_config_read(struct file *f, char __user *buf, /* rev==4 APU flag */ config[no_regs++] = adev->flags & AMD_IS_APU ? 1 : 0; + /* rev==5 CG flag upper 32bit */ + config[no_regs++] = upper_32_bits(adev->cg_flags); + while (size && (*pos < no_regs * 4)) { uint32_t value; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 49f734137f15..fb69b4188c8a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -1703,7 +1703,7 @@ int amdgpu_device_ip_set_powergating_state(void *dev, * clockgating is enabled. */ void amdgpu_device_ip_get_clockgating_state(struct amdgpu_device *adev, - u32 *flags) + u64 *flags) { int i; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_df.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_df.h index 6b25837955c4..1538b2dbfff1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_df.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_df.h @@ -40,7 +40,7 @@ struct amdgpu_df_funcs { void (*update_medium_grain_clock_gating)(struct amdgpu_device *adev, bool enable); void (*get_clockgating_state)(struct amdgpu_device *adev, - u32 *flags); + u64 *flags); void (*enable_ecc_force_par_wr_rmw)(struct amdgpu_device *adev, bool enable); int (*pmc_start)(struct amdgpu_device *adev, uint64_t config, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 29e9419a914b..4efaa183abcd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -136,7 +136,7 @@ int amdgpu_sched_jobs = 32; int amdgpu_sched_hw_submission = 2; uint amdgpu_pcie_gen_cap; uint amdgpu_pcie_lane_cap; -uint amdgpu_cg_mask = 0xffffffff; +u64 amdgpu_cg_mask = 0xffffffffffffffff; uint amdgpu_pg_mask = 0xffffffff; uint amdgpu_sdma_phase_quantum = 32; char *amdgpu_disable_cu = NULL; @@ -454,12 +454,12 @@ MODULE_PARM_DESC(pcie_lane_cap, "PCIE Lane Caps (0: autodetect (default))"); module_param_named(pcie_lane_cap, amdgpu_pcie_lane_cap, uint, 0444); /** - * DOC: cg_mask (uint) + * DOC: cg_mask (ullong) * Override Clockgating features enabled on GPU (0 = disable clock gating). See the AMD_CG_SUPPORT flags in - * drivers/gpu/drm/amd/include/amd_shared.h. The default is 0xffffffff (all enabled). + * drivers/gpu/drm/amd/include/amd_shared.h. The default is 0xffffffffffffffff (all enabled). */ MODULE_PARM_DESC(cg_mask, "Clockgating flags mask (0 = disable clock gating)"); -module_param_named(cg_mask, amdgpu_cg_mask, uint, 0444); +module_param_named(cg_mask, amdgpu_cg_mask, ullong, 0444); /** * DOC: pg_mask (uint) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.h index 9181c7bef7c6..ac5c61d3de2b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_hdp.h @@ -33,7 +33,7 @@ struct amdgpu_hdp_funcs { void (*invalidate_hdp)(struct amdgpu_device *adev, struct amdgpu_ring *ring); void (*update_clock_gating)(struct amdgpu_device *adev, bool enable); - void (*get_clock_gating_state)(struct amdgpu_device *adev, u32 *flags); + void (*get_clock_gating_state)(struct amdgpu_device *adev, u64 *flags); void (*init_registers)(struct amdgpu_device *adev); }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h index 9f1540f0ebf9..f939395c5914 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mmhub.h @@ -34,7 +34,7 @@ struct amdgpu_mmhub_funcs { void (*gart_disable)(struct amdgpu_device *adev); int (*set_clockgating)(struct amdgpu_device *adev, enum amd_clockgating_state state); - void (*get_clockgating)(struct amdgpu_device *adev, u32 *flags); + void (*get_clockgating)(struct amdgpu_device *adev, u64 *flags); void (*setup_vm_pt_regs)(struct amdgpu_device *adev, uint32_t vmid, uint64_t page_table_base); void (*update_power_gating)(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h index 3d13e601fc35..03439083182a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h @@ -83,7 +83,7 @@ struct amdgpu_nbio_funcs { void (*update_medium_grain_light_sleep)(struct amdgpu_device *adev, bool enable); void (*get_clockgating_state)(struct amdgpu_device *adev, - u32 *flags); + u64 *flags); void (*ih_control)(struct amdgpu_device *adev); void (*init_registers)(struct amdgpu_device *adev); void (*remap_hdp_registers)(struct amdgpu_device *adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_smuio.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_smuio.h index 484bb3dcec47..c7a823f3f2c5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_smuio.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_smuio.h @@ -27,7 +27,7 @@ struct amdgpu_smuio_funcs { u32 (*get_rom_index_offset)(struct amdgpu_device *adev); u32 (*get_rom_data_offset)(struct amdgpu_device *adev); void (*update_rom_clock_gating)(struct amdgpu_device *adev, bool enable); - void (*get_clock_gating_state)(struct amdgpu_device *adev, u32 *flags); + void (*get_clock_gating_state)(struct amdgpu_device *adev, u64 *flags); u32 (*get_die_id)(struct amdgpu_device *adev); u32 (*get_socket_id)(struct amdgpu_device *adev); bool (*is_host_gpu_xgmi_supported)(struct amdgpu_device *adev); diff --git a/drivers/gpu/drm/amd/amdgpu/athub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/athub_v1_0.c index 88642e7ecdf4..a13c443ea10f 100644 --- a/drivers/gpu/drm/amd/amdgpu/athub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/athub_v1_0.c @@ -87,7 +87,7 @@ int athub_v1_0_set_clockgating(struct amdgpu_device *adev, return 0; } -void athub_v1_0_get_clockgating(struct amdgpu_device *adev, u32 *flags) +void athub_v1_0_get_clockgating(struct amdgpu_device *adev, u64 *flags) { int data; diff --git a/drivers/gpu/drm/amd/amdgpu/athub_v1_0.h b/drivers/gpu/drm/amd/amdgpu/athub_v1_0.h index b279af59e34f..6be0a6704ea7 100644 --- a/drivers/gpu/drm/amd/amdgpu/athub_v1_0.h +++ b/drivers/gpu/drm/amd/amdgpu/athub_v1_0.h @@ -25,6 +25,6 @@ int athub_v1_0_set_clockgating(struct amdgpu_device *adev, enum amd_clockgating_state state); -void athub_v1_0_get_clockgating(struct amdgpu_device *adev, u32 *flags); +void athub_v1_0_get_clockgating(struct amdgpu_device *adev, u64 *flags); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/athub_v2_0.c b/drivers/gpu/drm/amd/amdgpu/athub_v2_0.c index a720436857b4..a9521c98e7f7 100644 --- a/drivers/gpu/drm/amd/amdgpu/athub_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/athub_v2_0.c @@ -93,7 +93,7 @@ int athub_v2_0_set_clockgating(struct amdgpu_device *adev, return 0; } -void athub_v2_0_get_clockgating(struct amdgpu_device *adev, u32 *flags) +void athub_v2_0_get_clockgating(struct amdgpu_device *adev, u64 *flags) { int data; diff --git a/drivers/gpu/drm/amd/amdgpu/athub_v2_0.h b/drivers/gpu/drm/amd/amdgpu/athub_v2_0.h index 02932c1c8bab..8b763f6dfd81 100644 --- a/drivers/gpu/drm/amd/amdgpu/athub_v2_0.h +++ b/drivers/gpu/drm/amd/amdgpu/athub_v2_0.h @@ -25,6 +25,6 @@ int athub_v2_0_set_clockgating(struct amdgpu_device *adev, enum amd_clockgating_state state); -void athub_v2_0_get_clockgating(struct amdgpu_device *adev, u32 *flags); +void athub_v2_0_get_clockgating(struct amdgpu_device *adev, u64 *flags); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/athub_v2_1.c b/drivers/gpu/drm/amd/amdgpu/athub_v2_1.c index ad8e87d3d2cb..78508ae6a670 100644 --- a/drivers/gpu/drm/amd/amdgpu/athub_v2_1.c +++ b/drivers/gpu/drm/amd/amdgpu/athub_v2_1.c @@ -85,7 +85,7 @@ int athub_v2_1_set_clockgating(struct amdgpu_device *adev, return 0; } -void athub_v2_1_get_clockgating(struct amdgpu_device *adev, u32 *flags) +void athub_v2_1_get_clockgating(struct amdgpu_device *adev, u64 *flags) { int data; diff --git a/drivers/gpu/drm/amd/amdgpu/athub_v2_1.h b/drivers/gpu/drm/amd/amdgpu/athub_v2_1.h index 5e6824c0f591..b799f14bce03 100644 --- a/drivers/gpu/drm/amd/amdgpu/athub_v2_1.h +++ b/drivers/gpu/drm/amd/amdgpu/athub_v2_1.h @@ -25,6 +25,6 @@ int athub_v2_1_set_clockgating(struct amdgpu_device *adev, enum amd_clockgating_state state); -void athub_v2_1_get_clockgating(struct amdgpu_device *adev, u32 *flags); +void athub_v2_1_get_clockgating(struct amdgpu_device *adev, u64 *flags); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/df_v1_7.c b/drivers/gpu/drm/amd/amdgpu/df_v1_7.c index 2d01ac0d4c11..b991609f46c1 100644 --- a/drivers/gpu/drm/amd/amdgpu/df_v1_7.c +++ b/drivers/gpu/drm/amd/amdgpu/df_v1_7.c @@ -99,7 +99,7 @@ static void df_v1_7_update_medium_grain_clock_gating(struct amdgpu_device *adev, } static void df_v1_7_get_clockgating_state(struct amdgpu_device *adev, - u32 *flags) + u64 *flags) { u32 tmp; diff --git a/drivers/gpu/drm/amd/amdgpu/df_v3_6.c b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c index f4dfca013ec5..483a441b46aa 100644 --- a/drivers/gpu/drm/amd/amdgpu/df_v3_6.c +++ b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c @@ -332,7 +332,7 @@ static void df_v3_6_update_medium_grain_clock_gating(struct amdgpu_device *adev, } static void df_v3_6_get_clockgating_state(struct amdgpu_device *adev, - u32 *flags) + u64 *flags) { u32 tmp; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index 9426e252d8aa..5147358f2400 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -8451,7 +8451,7 @@ static int gfx_v10_0_set_clockgating_state(void *handle, return 0; } -static void gfx_v10_0_get_clockgating_state(void *handle, u32 *flags) +static void gfx_v10_0_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 5f112efda634..46d436be74eb 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -5475,7 +5475,7 @@ static int gfx_v8_0_set_powergating_state(void *handle, return 0; } -static void gfx_v8_0_get_clockgating_state(void *handle, u32 *flags) +static void gfx_v8_0_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index b8cfcc6b1125..26535ebec498 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -5233,7 +5233,7 @@ static int gfx_v9_0_set_clockgating_state(void *handle, return 0; } -static void gfx_v9_0_get_clockgating_state(void *handle, u32 *flags) +static void gfx_v9_0_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c index 5228421b0f72..a455e59f41f4 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c @@ -1161,7 +1161,7 @@ static int gmc_v10_0_set_clockgating_state(void *handle, return athub_v2_0_set_clockgating(adev, state); } -static void gmc_v10_0_get_clockgating_state(void *handle, u32 *flags) +static void gmc_v10_0_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index 1932a3e4af7e..382dde1ce74c 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -1690,7 +1690,7 @@ static int gmc_v8_0_set_powergating_state(void *handle, return 0; } -static void gmc_v8_0_get_clockgating_state(void *handle, u32 *flags) +static void gmc_v8_0_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 6009fbfdcc19..22761a3bb818 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -1948,7 +1948,7 @@ static int gmc_v9_0_set_clockgating_state(void *handle, return 0; } -static void gmc_v9_0_get_clockgating_state(void *handle, u32 *flags) +static void gmc_v9_0_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; diff --git a/drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c b/drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c index 046216635262..adf89680f53e 100644 --- a/drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c @@ -124,7 +124,7 @@ static void hdp_v4_0_update_clock_gating(struct amdgpu_device *adev, } static void hdp_v4_0_get_clockgating_state(struct amdgpu_device *adev, - u32 *flags) + u64 *flags) { int data; diff --git a/drivers/gpu/drm/amd/amdgpu/hdp_v5_0.c b/drivers/gpu/drm/amd/amdgpu/hdp_v5_0.c index 5793977953cc..a9ea23fa0def 100644 --- a/drivers/gpu/drm/amd/amdgpu/hdp_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/hdp_v5_0.c @@ -181,7 +181,7 @@ static void hdp_v5_0_update_clock_gating(struct amdgpu_device *adev, } static void hdp_v5_0_get_clockgating_state(struct amdgpu_device *adev, - u32 *flags) + u64 *flags) { uint32_t tmp; diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c index 4c9f0c0f3116..3f44a099c52a 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c @@ -546,7 +546,7 @@ static int mmhub_v1_0_set_clockgating(struct amdgpu_device *adev, return 0; } -static void mmhub_v1_0_get_clockgating(struct amdgpu_device *adev, u32 *flags) +static void mmhub_v1_0_get_clockgating(struct amdgpu_device *adev, u64 *flags) { int data, data1; diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c index 3b901f941627..6fa7090bc6cb 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_7.c @@ -542,7 +542,7 @@ static int mmhub_v1_7_set_clockgating(struct amdgpu_device *adev, return 0; } -static void mmhub_v1_7_get_clockgating(struct amdgpu_device *adev, u32 *flags) +static void mmhub_v1_7_get_clockgating(struct amdgpu_device *adev, u64 *flags) { int data, data1; diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c index 3718ff610ab2..636abd855686 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c @@ -682,7 +682,7 @@ static int mmhub_v2_0_set_clockgating(struct amdgpu_device *adev, return 0; } -static void mmhub_v2_0_get_clockgating(struct amdgpu_device *adev, u32 *flags) +static void mmhub_v2_0_get_clockgating(struct amdgpu_device *adev, u64 *flags) { int data, data1; diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_3.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_3.c index 1957fb098c4d..ff44c5364a8c 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_3.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_3.c @@ -577,7 +577,7 @@ static int mmhub_v2_3_set_clockgating(struct amdgpu_device *adev, return 0; } -static void mmhub_v2_3_get_clockgating(struct amdgpu_device *adev, u32 *flags) +static void mmhub_v2_3_get_clockgating(struct amdgpu_device *adev, u64 *flags) { int data, data1, data2, data3; diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v9_4.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v9_4.c index 619106f7d23d..6e0145b2b408 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v9_4.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v9_4.c @@ -647,7 +647,7 @@ static int mmhub_v9_4_set_clockgating(struct amdgpu_device *adev, return 0; } -static void mmhub_v9_4_get_clockgating(struct amdgpu_device *adev, u32 *flags) +static void mmhub_v9_4_get_clockgating(struct amdgpu_device *adev, u64 *flags) { int data, data1; diff --git a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c index 8ce5b8ca1fd7..97201ab0965e 100644 --- a/drivers/gpu/drm/amd/amdgpu/navi10_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/navi10_ih.c @@ -685,7 +685,7 @@ static int navi10_ih_set_powergating_state(void *handle, return 0; } -static void navi10_ih_get_clockgating_state(void *handle, u32 *flags) +static void navi10_ih_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c b/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c index ee7cab37dfd5..6cd1fb2eb913 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c @@ -278,7 +278,7 @@ static void nbio_v2_3_update_medium_grain_light_sleep(struct amdgpu_device *adev } static void nbio_v2_3_get_clockgating_state(struct amdgpu_device *adev, - u32 *flags) + u64 *flags) { int data; diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c index 4bbacf1be25a..f7f6ddebd3e4 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c @@ -210,7 +210,7 @@ static void nbio_v6_1_update_medium_grain_light_sleep(struct amdgpu_device *adev } static void nbio_v6_1_get_clockgating_state(struct amdgpu_device *adev, - u32 *flags) + u64 *flags) { int data; diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c index 37a4039fdfc5..aa0326d00c72 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c @@ -205,7 +205,7 @@ static void nbio_v7_0_update_medium_grain_light_sleep(struct amdgpu_device *adev } static void nbio_v7_0_get_clockgating_state(struct amdgpu_device *adev, - u32 *flags) + u64 *flags) { int data; diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c index 6f81de6f3cc4..31776b12e4c4 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_2.c @@ -306,7 +306,7 @@ static void nbio_v7_2_update_medium_grain_light_sleep(struct amdgpu_device *adev } static void nbio_v7_2_get_clockgating_state(struct amdgpu_device *adev, - u32 *flags) + u64 *flags) { int data; diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c index c2357e83a8c4..4531761dcf77 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c @@ -273,7 +273,7 @@ static void nbio_v7_4_update_medium_grain_light_sleep(struct amdgpu_device *adev } static void nbio_v7_4_get_clockgating_state(struct amdgpu_device *adev, - u32 *flags) + u64 *flags) { int data; diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c index e19f14c3ef59..0a7946c59a42 100644 --- a/drivers/gpu/drm/amd/amdgpu/nv.c +++ b/drivers/gpu/drm/amd/amdgpu/nv.c @@ -1115,7 +1115,7 @@ static int nv_common_set_powergating_state(void *handle, return 0; } -static void nv_common_get_clockgating_state(void *handle, u32 *flags) +static void nv_common_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c index 4ef4feff5649..3695374896ed 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c @@ -1535,7 +1535,7 @@ static int sdma_v3_0_set_powergating_state(void *handle, return 0; } -static void sdma_v3_0_get_clockgating_state(void *handle, u32 *flags) +static void sdma_v3_0_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c index d7e8f7232364..8589ab1c9800 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c @@ -2372,7 +2372,7 @@ static int sdma_v4_0_set_powergating_state(void *handle, return 0; } -static void sdma_v4_0_get_clockgating_state(void *handle, u32 *flags) +static void sdma_v4_0_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c index a8d49c005f73..775aabde1ae2 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c @@ -1648,7 +1648,7 @@ static int sdma_v5_0_set_powergating_state(void *handle, return 0; } -static void sdma_v5_0_get_clockgating_state(void *handle, u32 *flags) +static void sdma_v5_0_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c index 824eace69884..ca50857b982d 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c @@ -1645,7 +1645,7 @@ static int sdma_v5_2_set_powergating_state(void *handle, return 0; } -static void sdma_v5_2_get_clockgating_state(void *handle, u32 *flags) +static void sdma_v5_2_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/amdgpu/smuio_v11_0.c b/drivers/gpu/drm/amd/amdgpu/smuio_v11_0.c index b6f1322f908c..acdc40f99ab3 100644 --- a/drivers/gpu/drm/amd/amdgpu/smuio_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/smuio_v11_0.c @@ -59,7 +59,7 @@ static void smuio_v11_0_update_rom_clock_gating(struct amdgpu_device *adev, bool WREG32_SOC15(SMUIO, 0, mmCGTT_ROM_CLK_CTRL0, data); } -static void smuio_v11_0_get_clock_gating_state(struct amdgpu_device *adev, u32 *flags) +static void smuio_v11_0_get_clock_gating_state(struct amdgpu_device *adev, u64 *flags) { u32 data; diff --git a/drivers/gpu/drm/amd/amdgpu/smuio_v11_0_6.c b/drivers/gpu/drm/amd/amdgpu/smuio_v11_0_6.c index 3a18dbb55c32..2afeb8b37f62 100644 --- a/drivers/gpu/drm/amd/amdgpu/smuio_v11_0_6.c +++ b/drivers/gpu/drm/amd/amdgpu/smuio_v11_0_6.c @@ -56,7 +56,7 @@ static void smuio_v11_0_6_update_rom_clock_gating(struct amdgpu_device *adev, bo WREG32_SOC15(SMUIO, 0, mmCGTT_ROM_CLK_CTRL0, data); } -static void smuio_v11_0_6_get_clock_gating_state(struct amdgpu_device *adev, u32 *flags) +static void smuio_v11_0_6_get_clock_gating_state(struct amdgpu_device *adev, u64 *flags) { u32 data; diff --git a/drivers/gpu/drm/amd/amdgpu/smuio_v13_0.c b/drivers/gpu/drm/amd/amdgpu/smuio_v13_0.c index 39b7c206770f..13e905c22592 100644 --- a/drivers/gpu/drm/amd/amdgpu/smuio_v13_0.c +++ b/drivers/gpu/drm/amd/amdgpu/smuio_v13_0.c @@ -58,7 +58,7 @@ static void smuio_v13_0_update_rom_clock_gating(struct amdgpu_device *adev, bool WREG32_SOC15(SMUIO, 0, regCGTT_ROM_CLK_CTRL0, data); } -static void smuio_v13_0_get_clock_gating_state(struct amdgpu_device *adev, u32 *flags) +static void smuio_v13_0_get_clock_gating_state(struct amdgpu_device *adev, u64 *flags) { u32 data; diff --git a/drivers/gpu/drm/amd/amdgpu/smuio_v9_0.c b/drivers/gpu/drm/amd/amdgpu/smuio_v9_0.c index 8417890af227..e4e30b9d481b 100644 --- a/drivers/gpu/drm/amd/amdgpu/smuio_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/smuio_v9_0.c @@ -56,7 +56,7 @@ static void smuio_v9_0_update_rom_clock_gating(struct amdgpu_device *adev, bool WREG32_SOC15(SMUIO, 0, mmCGTT_ROM_CLK_CTRL0, data); } -static void smuio_v9_0_get_clock_gating_state(struct amdgpu_device *adev, u32 *flags) +static void smuio_v9_0_get_clock_gating_state(struct amdgpu_device *adev, u64 *flags) { u32 data; diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index 3d0251ef8d79..3ee7322081d2 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -1419,7 +1419,7 @@ static int soc15_common_set_clockgating_state(void *handle, return 0; } -static void soc15_common_get_clockgating_state(void *handle, u32 *flags) +static void soc15_common_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c index 563493d1f830..d7e31e48a2b8 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c @@ -833,7 +833,7 @@ out: return ret; } -static void uvd_v5_0_get_clockgating_state(void *handle, u32 *flags) +static void uvd_v5_0_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c index 2d558c2f417d..375c440957dc 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c @@ -1494,7 +1494,7 @@ out: return ret; } -static void uvd_v6_0_get_clockgating_state(void *handle, u32 *flags) +static void uvd_v6_0_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c index 142e291983b4..8def62c83ffd 100644 --- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c @@ -831,7 +831,7 @@ out: return ret; } -static void vce_v3_0_get_clockgating_state(void *handle, u32 *flags) +static void vce_v3_0_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c index 039b90cdc3bc..c5b88d15a6df 100644 --- a/drivers/gpu/drm/amd/amdgpu/vi.c +++ b/drivers/gpu/drm/amd/amdgpu/vi.c @@ -2033,7 +2033,7 @@ static int vi_common_set_powergating_state(void *handle, return 0; } -static void vi_common_get_clockgating_state(void *handle, u32 *flags) +static void vi_common_get_clockgating_state(void *handle, u64 *flags) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; int data; diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h index fe4e585781bb..741dae17562a 100644 --- a/drivers/gpu/drm/amd/include/amd_shared.h +++ b/drivers/gpu/drm/amd/include/amd_shared.h @@ -116,38 +116,38 @@ enum amd_powergating_state { /* CG flags */ -#define AMD_CG_SUPPORT_GFX_MGCG (1 << 0) -#define AMD_CG_SUPPORT_GFX_MGLS (1 << 1) -#define AMD_CG_SUPPORT_GFX_CGCG (1 << 2) -#define AMD_CG_SUPPORT_GFX_CGLS (1 << 3) -#define AMD_CG_SUPPORT_GFX_CGTS (1 << 4) -#define AMD_CG_SUPPORT_GFX_CGTS_LS (1 << 5) -#define AMD_CG_SUPPORT_GFX_CP_LS (1 << 6) -#define AMD_CG_SUPPORT_GFX_RLC_LS (1 << 7) -#define AMD_CG_SUPPORT_MC_LS (1 << 8) -#define AMD_CG_SUPPORT_MC_MGCG (1 << 9) -#define AMD_CG_SUPPORT_SDMA_LS (1 << 10) -#define AMD_CG_SUPPORT_SDMA_MGCG (1 << 11) -#define AMD_CG_SUPPORT_BIF_LS (1 << 12) -#define AMD_CG_SUPPORT_UVD_MGCG (1 << 13) -#define AMD_CG_SUPPORT_VCE_MGCG (1 << 14) -#define AMD_CG_SUPPORT_HDP_LS (1 << 15) -#define AMD_CG_SUPPORT_HDP_MGCG (1 << 16) -#define AMD_CG_SUPPORT_ROM_MGCG (1 << 17) -#define AMD_CG_SUPPORT_DRM_LS (1 << 18) -#define AMD_CG_SUPPORT_BIF_MGCG (1 << 19) -#define AMD_CG_SUPPORT_GFX_3D_CGCG (1 << 20) -#define AMD_CG_SUPPORT_GFX_3D_CGLS (1 << 21) -#define AMD_CG_SUPPORT_DRM_MGCG (1 << 22) -#define AMD_CG_SUPPORT_DF_MGCG (1 << 23) -#define AMD_CG_SUPPORT_VCN_MGCG (1 << 24) -#define AMD_CG_SUPPORT_HDP_DS (1 << 25) -#define AMD_CG_SUPPORT_HDP_SD (1 << 26) -#define AMD_CG_SUPPORT_IH_CG (1 << 27) -#define AMD_CG_SUPPORT_ATHUB_LS (1 << 28) -#define AMD_CG_SUPPORT_ATHUB_MGCG (1 << 29) -#define AMD_CG_SUPPORT_JPEG_MGCG (1 << 30) -#define AMD_CG_SUPPORT_GFX_FGCG (1 << 31) +#define AMD_CG_SUPPORT_GFX_MGCG (1ULL << 0) +#define AMD_CG_SUPPORT_GFX_MGLS (1ULL << 1) +#define AMD_CG_SUPPORT_GFX_CGCG (1ULL << 2) +#define AMD_CG_SUPPORT_GFX_CGLS (1ULL << 3) +#define AMD_CG_SUPPORT_GFX_CGTS (1ULL << 4) +#define AMD_CG_SUPPORT_GFX_CGTS_LS (1ULL << 5) +#define AMD_CG_SUPPORT_GFX_CP_LS (1ULL << 6) +#define AMD_CG_SUPPORT_GFX_RLC_LS (1ULL << 7) +#define AMD_CG_SUPPORT_MC_LS (1ULL << 8) +#define AMD_CG_SUPPORT_MC_MGCG (1ULL << 9) +#define AMD_CG_SUPPORT_SDMA_LS (1ULL << 10) +#define AMD_CG_SUPPORT_SDMA_MGCG (1ULL << 11) +#define AMD_CG_SUPPORT_BIF_LS (1ULL << 12) +#define AMD_CG_SUPPORT_UVD_MGCG (1ULL << 13) +#define AMD_CG_SUPPORT_VCE_MGCG (1ULL << 14) +#define AMD_CG_SUPPORT_HDP_LS (1ULL << 15) +#define AMD_CG_SUPPORT_HDP_MGCG (1ULL << 16) +#define AMD_CG_SUPPORT_ROM_MGCG (1ULL << 17) +#define AMD_CG_SUPPORT_DRM_LS (1ULL << 18) +#define AMD_CG_SUPPORT_BIF_MGCG (1ULL << 19) +#define AMD_CG_SUPPORT_GFX_3D_CGCG (1ULL << 20) +#define AMD_CG_SUPPORT_GFX_3D_CGLS (1ULL << 21) +#define AMD_CG_SUPPORT_DRM_MGCG (1ULL << 22) +#define AMD_CG_SUPPORT_DF_MGCG (1ULL << 23) +#define AMD_CG_SUPPORT_VCN_MGCG (1ULL << 24) +#define AMD_CG_SUPPORT_HDP_DS (1ULL << 25) +#define AMD_CG_SUPPORT_HDP_SD (1ULL << 26) +#define AMD_CG_SUPPORT_IH_CG (1ULL << 27) +#define AMD_CG_SUPPORT_ATHUB_LS (1ULL << 28) +#define AMD_CG_SUPPORT_ATHUB_MGCG (1ULL << 29) +#define AMD_CG_SUPPORT_JPEG_MGCG (1ULL << 30) +#define AMD_CG_SUPPORT_GFX_FGCG (1ULL << 31) /* PG flags */ #define AMD_PG_SUPPORT_GFX_PG (1 << 0) #define AMD_PG_SUPPORT_GFX_SMG (1 << 1) @@ -298,7 +298,7 @@ struct amd_ip_funcs { enum amd_clockgating_state state); int (*set_powergating_state)(void *handle, enum amd_powergating_state state); - void (*get_clockgating_state)(void *handle, u32 *flags); + void (*get_clockgating_state)(void *handle, u64 *flags); }; diff --git a/drivers/gpu/drm/amd/pm/amdgpu_pm.c b/drivers/gpu/drm/amd/pm/amdgpu_pm.c index ec4ea141bfda..78ec9b71197d 100644 --- a/drivers/gpu/drm/amd/pm/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/pm/amdgpu_pm.c @@ -3526,7 +3526,7 @@ static int amdgpu_debugfs_pm_info_pp(struct seq_file *m, struct amdgpu_device *a return 0; } -static void amdgpu_parse_cg_state(struct seq_file *m, u32 flags) +static void amdgpu_parse_cg_state(struct seq_file *m, u64 flags) { int i; @@ -3539,7 +3539,7 @@ static int amdgpu_debugfs_pm_info_show(struct seq_file *m, void *unused) { struct amdgpu_device *adev = (struct amdgpu_device *)m->private; struct drm_device *dev = adev_to_drm(adev); - u32 flags = 0; + u64 flags = 0; int r; if (amdgpu_in_reset(adev)) @@ -3561,7 +3561,7 @@ static int amdgpu_debugfs_pm_info_show(struct seq_file *m, void *unused) amdgpu_device_ip_get_clockgating_state(adev, &flags); - seq_printf(m, "Clock Gating Flags Mask: 0x%x\n", flags); + seq_printf(m, "Clock Gating Flags Mask: 0x%llx\n", flags); amdgpu_parse_cg_state(m, flags); seq_printf(m, "\n"); diff --git a/drivers/gpu/drm/amd/pm/inc/amdgpu_pm.h b/drivers/gpu/drm/amd/pm/inc/amdgpu_pm.h index a920515e2274..52045ad59bed 100644 --- a/drivers/gpu/drm/amd/pm/inc/amdgpu_pm.h +++ b/drivers/gpu/drm/amd/pm/inc/amdgpu_pm.h @@ -26,7 +26,7 @@ struct cg_flag_name { - u32 flag; + u64 flag; const char *name; }; -- cgit v1.2.3 From 73bce7a423c3bb7c3441177aca868ff5f2cf93fa Mon Sep 17 00:00:00 2001 From: Lijo Lazar Date: Fri, 8 Apr 2022 09:11:04 +0530 Subject: drm/amdgpu: Use flexible array member MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Use flexible array member in ip discovery struct as recommended[1]. [1] https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays v2: squash in struct_size fixes Signed-off-by: Lijo Lazar Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c | 8 ++++---- drivers/gpu/drm/amd/include/discovery.h | 2 +- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c index 5a182288391b..aaf2fc6b1a82 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c @@ -430,7 +430,7 @@ static void amdgpu_discovery_read_harvest_bit_per_ip(struct amdgpu_device *adev, } } next_ip: - ip_offset += sizeof(*ip) + 4 * (ip->num_base_address - 1); + ip_offset += struct_size(ip, base_address, ip->num_base_address); } } } @@ -798,7 +798,7 @@ static int amdgpu_discovery_sysfs_ips(struct amdgpu_device *adev, res = kobject_add(&ip_hw_instance->kobj, NULL, "%d", ip_hw_instance->num_instance); next_ip: - ip_offset += sizeof(*ip) + 4 * (ip->num_base_address - 1); + ip_offset += struct_size(ip, base_address, ip->num_base_address); } } @@ -1063,7 +1063,7 @@ int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev) } next_ip: - ip_offset += sizeof(*ip) + 4 * (ip->num_base_address - 1); + ip_offset += struct_size(ip, base_address, ip->num_base_address); } } @@ -1113,7 +1113,7 @@ int amdgpu_discovery_get_ip_version(struct amdgpu_device *adev, int hw_id, int n *revision = ip->revision; return 0; } - ip_offset += sizeof(*ip) + 4 * (ip->num_base_address - 1); + ip_offset += struct_size(ip, base_address, ip->num_base_address); } } diff --git a/drivers/gpu/drm/amd/include/discovery.h b/drivers/gpu/drm/amd/include/discovery.h index a486769b66c6..b25026c3ec96 100644 --- a/drivers/gpu/drm/amd/include/discovery.h +++ b/drivers/gpu/drm/amd/include/discovery.h @@ -93,7 +93,7 @@ typedef struct ip uint8_t harvest : 4; /* Harvest */ uint8_t reserved : 4; /* Placeholder field */ #endif - uint32_t base_address[1]; /* variable number of Addresses */ + uint32_t base_address[]; /* variable number of Addresses */ } ip; typedef struct die_header -- cgit v1.2.3 From edd11922e3d01549fc6f5a1fa1413eb2f7d5be55 Mon Sep 17 00:00:00 2001 From: Philip Yang Date: Tue, 5 Apr 2022 22:24:42 -0400 Subject: drm/amdkfd: Handle drain retry fault race with XNACK mode change Application could change XNACK enabled to disabled while KFD is draining stale retry fault, therefore the check for whether to drain retry faults must be before the check for whether xnack_enabled, to avoid report incorrect vm fault after application changes XNACK mode. Signed-off-by: Philip Yang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index d3fb2d0b5a25..8b6adc142e25 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -2685,11 +2685,6 @@ svm_range_restore_pages(struct amdgpu_device *adev, unsigned int pasid, pr_debug("kfd process not founded pasid 0x%x\n", pasid); return 0; } - if (!p->xnack_enabled) { - pr_debug("XNACK not enabled for pasid 0x%x\n", pasid); - r = -EFAULT; - goto out; - } svms = &p->svms; pr_debug("restoring svms 0x%p fault address 0x%llx\n", svms, addr); @@ -2700,6 +2695,12 @@ svm_range_restore_pages(struct amdgpu_device *adev, unsigned int pasid, goto out; } + if (!p->xnack_enabled) { + pr_debug("XNACK not enabled for pasid 0x%x\n", pasid); + r = -EFAULT; + goto out; + } + /* p->lead_thread is available as kfd_process_wq_release flush the work * before releasing task ref. */ -- cgit v1.2.3 From d9e50239a9611b9a1759e007e9a810c8d178da28 Mon Sep 17 00:00:00 2001 From: Yongqiang Sun Date: Fri, 8 Apr 2022 09:26:14 -0400 Subject: drm/amd/amdgpu: Fix asm/hypervisor.h build error. Add CONFIG_X86 check to fix the build error. Fixes: 49aa98ca30cd18 ("drm/amd/amdgpu: Only reserve vram for firmware with vega9 MS_HYPERV host.") Reported-by: kernel test robot Signed-off-by: Yongqiang Sun Acked-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c index 3e9582c245bb..88b852b3a2cb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c @@ -25,7 +25,9 @@ */ #include +#ifdef CONFIG_X86 #include +#endif #include "amdgpu.h" #include "amdgpu_gmc.h" @@ -650,10 +652,12 @@ void amdgpu_gmc_get_vbios_allocations(struct amdgpu_device *adev) /* * VEGA10 SRIOV VF with MS_HYPERV host needs some firmware reserved area. */ +#ifdef CONFIG_X86 if (amdgpu_sriov_vf(adev) && hypervisor_is_type(X86_HYPER_MS_HYPERV)) { adev->mman.stolen_reserved_offset = 0x500000; adev->mman.stolen_reserved_size = 0x200000; } +#endif break; case CHIP_RAVEN: case CHIP_RENOIR: -- cgit v1.2.3 From dc2947b35f519efbcdd67815ec965b8c9056d274 Mon Sep 17 00:00:00 2001 From: Tom St Denis Date: Fri, 8 Apr 2022 09:14:22 -0400 Subject: drm/amd/amdgpu: Update debugfs GCA data The data revision was not changed to 5 from 4 when the CG flags were extended to 64-bits. Since this was missed I took the opportunity to add future upper 64-bits of PG flags as well so we don't need to bump it again when that comes. Signed-off-by: Tom St Denis Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c index 1db48dc58ae9..13e4d8f9b874 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c @@ -730,7 +730,7 @@ static ssize_t amdgpu_debugfs_gca_config_read(struct file *f, char __user *buf, return -ENOMEM; /* version, increment each time something is added */ - config[no_regs++] = 4; + config[no_regs++] = 5; config[no_regs++] = adev->gfx.config.max_shader_engines; config[no_regs++] = adev->gfx.config.max_tile_pipes; config[no_regs++] = adev->gfx.config.max_cu_per_sh; @@ -757,7 +757,7 @@ static ssize_t amdgpu_debugfs_gca_config_read(struct file *f, char __user *buf, /* rev==1 */ config[no_regs++] = adev->rev_id; - config[no_regs++] = adev->pg_flags; + config[no_regs++] = lower_32_bits(adev->pg_flags); config[no_regs++] = lower_32_bits(adev->cg_flags); /* rev==2 */ @@ -773,7 +773,8 @@ static ssize_t amdgpu_debugfs_gca_config_read(struct file *f, char __user *buf, /* rev==4 APU flag */ config[no_regs++] = adev->flags & AMD_IS_APU ? 1 : 0; - /* rev==5 CG flag upper 32bit */ + /* rev==5 PG/CG flag upper 32bit */ + config[no_regs++] = upper_32_bits(adev->pg_flags); config[no_regs++] = upper_32_bits(adev->cg_flags); while (size && (*pos < no_regs * 4)) { -- cgit v1.2.3 From 024106930ce7d8192f97f0341f233e8d34db7e5c Mon Sep 17 00:00:00 2001 From: Tom Rix Date: Thu, 7 Apr 2022 17:46:59 -0400 Subject: drm/radeon: change cayman_default_state table from global to static cayman_default_state and cayman_default_size are only used in ni.c. Single file symbols should be static. So move their definitions to cayman_blit_shaders.h and change their storage-class-specifier to static. Remove unneeded cayman_blit_shader.c cayman_ps/vs definitions were removed with commit 4f8629675800 ("drm/radeon/kms: remove r6xx+ blit copy routines") So their declarations in cayman_blit_shader.h are not needed, so remove them. Signed-off-by: Tom Rix Signed-off-by: Alex Deucher --- drivers/gpu/drm/radeon/Makefile | 2 +- drivers/gpu/drm/radeon/cayman_blit_shaders.c | 320 --------------------------- drivers/gpu/drm/radeon/cayman_blit_shaders.h | 294 +++++++++++++++++++++++- 3 files changed, 290 insertions(+), 326 deletions(-) delete mode 100644 drivers/gpu/drm/radeon/cayman_blit_shaders.c (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile index 664381f4eb07..2425a3612d6c 100644 --- a/drivers/gpu/drm/radeon/Makefile +++ b/drivers/gpu/drm/radeon/Makefile @@ -42,7 +42,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \ r200.o radeon_legacy_tv.o r600_cs.o r600_blit_shaders.o \ radeon_pm.o atombios_dp.o r600_hdmi.o dce3_1_afmt.o \ evergreen.o evergreen_cs.o evergreen_blit_shaders.o \ - evergreen_hdmi.o radeon_trace_points.o ni.o cayman_blit_shaders.o \ + evergreen_hdmi.o radeon_trace_points.o ni.o \ atombios_encoders.o radeon_semaphore.o radeon_sa.o atombios_i2c.o si.o \ radeon_prime.o cik.o cik_blit_shaders.o \ r600_dpm.o rs780_dpm.o rv6xx_dpm.o rv770_dpm.o rv730_dpm.o rv740_dpm.o \ diff --git a/drivers/gpu/drm/radeon/cayman_blit_shaders.c b/drivers/gpu/drm/radeon/cayman_blit_shaders.c deleted file mode 100644 index 9fec4d09f383..000000000000 --- a/drivers/gpu/drm/radeon/cayman_blit_shaders.c +++ /dev/null @@ -1,320 +0,0 @@ -/* - * Copyright 2010 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice (including the next - * paragraph) shall be included in all copies or substantial portions of the - * Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER - * DEALINGS IN THE SOFTWARE. - * - * Authors: - * Alex Deucher - */ - -#include -#include -#include - -/* - * evergreen cards need to use the 3D engine to blit data which requires - * quite a bit of hw state setup. Rather than pull the whole 3D driver - * (which normally generates the 3D state) into the DRM, we opt to use - * statically generated state tables. The register state and shaders - * were hand generated to support blitting functionality. See the 3D - * driver or documentation for descriptions of the registers and - * shader instructions. - */ - -const u32 cayman_default_state[] = -{ - 0xc0066900, - 0x00000000, - 0x00000060, /* DB_RENDER_CONTROL */ - 0x00000000, /* DB_COUNT_CONTROL */ - 0x00000000, /* DB_DEPTH_VIEW */ - 0x0000002a, /* DB_RENDER_OVERRIDE */ - 0x00000000, /* DB_RENDER_OVERRIDE2 */ - 0x00000000, /* DB_HTILE_DATA_BASE */ - - 0xc0026900, - 0x0000000a, - 0x00000000, /* DB_STENCIL_CLEAR */ - 0x00000000, /* DB_DEPTH_CLEAR */ - - 0xc0036900, - 0x0000000f, - 0x00000000, /* DB_DEPTH_INFO */ - 0x00000000, /* DB_Z_INFO */ - 0x00000000, /* DB_STENCIL_INFO */ - - 0xc0016900, - 0x00000080, - 0x00000000, /* PA_SC_WINDOW_OFFSET */ - - 0xc00d6900, - 0x00000083, - 0x0000ffff, /* PA_SC_CLIPRECT_RULE */ - 0x00000000, /* PA_SC_CLIPRECT_0_TL */ - 0x20002000, /* PA_SC_CLIPRECT_0_BR */ - 0x00000000, - 0x20002000, - 0x00000000, - 0x20002000, - 0x00000000, - 0x20002000, - 0xaaaaaaaa, /* PA_SC_EDGERULE */ - 0x00000000, /* PA_SU_HARDWARE_SCREEN_OFFSET */ - 0x0000000f, /* CB_TARGET_MASK */ - 0x0000000f, /* CB_SHADER_MASK */ - - 0xc0226900, - 0x00000094, - 0x80000000, /* PA_SC_VPORT_SCISSOR_0_TL */ - 0x20002000, /* PA_SC_VPORT_SCISSOR_0_BR */ - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x00000000, /* PA_SC_VPORT_ZMIN_0 */ - 0x3f800000, /* PA_SC_VPORT_ZMAX_0 */ - - 0xc0016900, - 0x000000d4, - 0x00000000, /* SX_MISC */ - - 0xc0026900, - 0x000000d9, - 0x00000000, /* CP_RINGID */ - 0x00000000, /* CP_VMID */ - - 0xc0096900, - 0x00000100, - 0x00ffffff, /* VGT_MAX_VTX_INDX */ - 0x00000000, /* VGT_MIN_VTX_INDX */ - 0x00000000, /* VGT_INDX_OFFSET */ - 0x00000000, /* VGT_MULTI_PRIM_IB_RESET_INDX */ - 0x00000000, /* SX_ALPHA_TEST_CONTROL */ - 0x00000000, /* CB_BLEND_RED */ - 0x00000000, /* CB_BLEND_GREEN */ - 0x00000000, /* CB_BLEND_BLUE */ - 0x00000000, /* CB_BLEND_ALPHA */ - - 0xc0016900, - 0x00000187, - 0x00000100, /* SPI_VS_OUT_ID_0 */ - - 0xc0026900, - 0x00000191, - 0x00000100, /* SPI_PS_INPUT_CNTL_0 */ - 0x00000101, /* SPI_PS_INPUT_CNTL_1 */ - - 0xc0016900, - 0x000001b1, - 0x00000000, /* SPI_VS_OUT_CONFIG */ - - 0xc0106900, - 0x000001b3, - 0x20000001, /* SPI_PS_IN_CONTROL_0 */ - 0x00000000, /* SPI_PS_IN_CONTROL_1 */ - 0x00000000, /* SPI_INTERP_CONTROL_0 */ - 0x00000000, /* SPI_INPUT_Z */ - 0x00000000, /* SPI_FOG_CNTL */ - 0x00100000, /* SPI_BARYC_CNTL */ - 0x00000000, /* SPI_PS_IN_CONTROL_2 */ - 0x00000000, /* SPI_COMPUTE_INPUT_CNTL */ - 0x00000000, /* SPI_COMPUTE_NUM_THREAD_X */ - 0x00000000, /* SPI_COMPUTE_NUM_THREAD_Y */ - 0x00000000, /* SPI_COMPUTE_NUM_THREAD_Z */ - 0x00000000, /* SPI_GPR_MGMT */ - 0x00000000, /* SPI_LDS_MGMT */ - 0x00000000, /* SPI_STACK_MGMT */ - 0x00000000, /* SPI_WAVE_MGMT_1 */ - 0x00000000, /* SPI_WAVE_MGMT_2 */ - - 0xc0016900, - 0x000001e0, - 0x00000000, /* CB_BLEND0_CONTROL */ - - 0xc00e6900, - 0x00000200, - 0x00000000, /* DB_DEPTH_CONTROL */ - 0x00000000, /* DB_EQAA */ - 0x00cc0010, /* CB_COLOR_CONTROL */ - 0x00000210, /* DB_SHADER_CONTROL */ - 0x00010000, /* PA_CL_CLIP_CNTL */ - 0x00000004, /* PA_SU_SC_MODE_CNTL */ - 0x00000100, /* PA_CL_VTE_CNTL */ - 0x00000000, /* PA_CL_VS_OUT_CNTL */ - 0x00000000, /* PA_CL_NANINF_CNTL */ - 0x00000000, /* PA_SU_LINE_STIPPLE_CNTL */ - 0x00000000, /* PA_SU_LINE_STIPPLE_SCALE */ - 0x00000000, /* PA_SU_PRIM_FILTER_CNTL */ - 0x00000000, /* */ - 0x00000000, /* */ - - 0xc0026900, - 0x00000229, - 0x00000000, /* SQ_PGM_START_FS */ - 0x00000000, - - 0xc0016900, - 0x0000023b, - 0x00000000, /* SQ_LDS_ALLOC_PS */ - - 0xc0066900, - 0x00000240, - 0x00000000, /* SQ_ESGS_RING_ITEMSIZE */ - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - - 0xc0046900, - 0x00000247, - 0x00000000, /* SQ_GS_VERT_ITEMSIZE */ - 0x00000000, - 0x00000000, - 0x00000000, - - 0xc0116900, - 0x00000280, - 0x00000000, /* PA_SU_POINT_SIZE */ - 0x00000000, /* PA_SU_POINT_MINMAX */ - 0x00000008, /* PA_SU_LINE_CNTL */ - 0x00000000, /* PA_SC_LINE_STIPPLE */ - 0x00000000, /* VGT_OUTPUT_PATH_CNTL */ - 0x00000000, /* VGT_HOS_CNTL */ - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, /* VGT_GS_MODE */ - - 0xc0026900, - 0x00000292, - 0x00000000, /* PA_SC_MODE_CNTL_0 */ - 0x00000000, /* PA_SC_MODE_CNTL_1 */ - - 0xc0016900, - 0x000002a1, - 0x00000000, /* VGT_PRIMITIVEID_EN */ - - 0xc0016900, - 0x000002a5, - 0x00000000, /* VGT_MULTI_PRIM_IB_RESET_EN */ - - 0xc0026900, - 0x000002a8, - 0x00000000, /* VGT_INSTANCE_STEP_RATE_0 */ - 0x00000000, - - 0xc0026900, - 0x000002ad, - 0x00000000, /* VGT_REUSE_OFF */ - 0x00000000, - - 0xc0016900, - 0x000002d5, - 0x00000000, /* VGT_SHADER_STAGES_EN */ - - 0xc0016900, - 0x000002dc, - 0x0000aa00, /* DB_ALPHA_TO_MASK */ - - 0xc0066900, - 0x000002de, - 0x00000000, /* PA_SU_POLY_OFFSET_DB_FMT_CNTL */ - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - - 0xc0026900, - 0x000002e5, - 0x00000000, /* VGT_STRMOUT_CONFIG */ - 0x00000000, - - 0xc01b6900, - 0x000002f5, - 0x76543210, /* PA_SC_CENTROID_PRIORITY_0 */ - 0xfedcba98, /* PA_SC_CENTROID_PRIORITY_1 */ - 0x00000000, /* PA_SC_LINE_CNTL */ - 0x00000000, /* PA_SC_AA_CONFIG */ - 0x00000005, /* PA_SU_VTX_CNTL */ - 0x3f800000, /* PA_CL_GB_VERT_CLIP_ADJ */ - 0x3f800000, /* PA_CL_GB_VERT_DISC_ADJ */ - 0x3f800000, /* PA_CL_GB_HORZ_CLIP_ADJ */ - 0x3f800000, /* PA_CL_GB_HORZ_DISC_ADJ */ - 0x00000000, /* PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_0 */ - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0xffffffff, /* PA_SC_AA_MASK_X0Y0_X1Y0 */ - 0xffffffff, - - 0xc0026900, - 0x00000316, - 0x0000000e, /* VGT_VERTEX_REUSE_BLOCK_CNTL */ - 0x00000010, /* */ -}; - -const u32 cayman_default_size = ARRAY_SIZE(cayman_default_state); diff --git a/drivers/gpu/drm/radeon/cayman_blit_shaders.h b/drivers/gpu/drm/radeon/cayman_blit_shaders.h index f5d0e9a60267..1dca73d9e005 100644 --- a/drivers/gpu/drm/radeon/cayman_blit_shaders.h +++ b/drivers/gpu/drm/radeon/cayman_blit_shaders.h @@ -20,16 +20,300 @@ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER * DEALINGS IN THE SOFTWARE. * + * Authors: + * Alex Deucher */ #ifndef CAYMAN_BLIT_SHADERS_H #define CAYMAN_BLIT_SHADERS_H -extern const u32 cayman_ps[]; -extern const u32 cayman_vs[]; -extern const u32 cayman_default_state[]; +/* + * evergreen cards need to use the 3D engine to blit data which requires + * quite a bit of hw state setup. Rather than pull the whole 3D driver + * (which normally generates the 3D state) into the DRM, we opt to use + * statically generated state tables. The register state and shaders + * were hand generated to support blitting functionality. See the 3D + * driver or documentation for descriptions of the registers and + * shader instructions. + */ +static const u32 cayman_default_state[] = { + 0xc0066900, + 0x00000000, + 0x00000060, /* DB_RENDER_CONTROL */ + 0x00000000, /* DB_COUNT_CONTROL */ + 0x00000000, /* DB_DEPTH_VIEW */ + 0x0000002a, /* DB_RENDER_OVERRIDE */ + 0x00000000, /* DB_RENDER_OVERRIDE2 */ + 0x00000000, /* DB_HTILE_DATA_BASE */ + + 0xc0026900, + 0x0000000a, + 0x00000000, /* DB_STENCIL_CLEAR */ + 0x00000000, /* DB_DEPTH_CLEAR */ + + 0xc0036900, + 0x0000000f, + 0x00000000, /* DB_DEPTH_INFO */ + 0x00000000, /* DB_Z_INFO */ + 0x00000000, /* DB_STENCIL_INFO */ + + 0xc0016900, + 0x00000080, + 0x00000000, /* PA_SC_WINDOW_OFFSET */ + + 0xc00d6900, + 0x00000083, + 0x0000ffff, /* PA_SC_CLIPRECT_RULE */ + 0x00000000, /* PA_SC_CLIPRECT_0_TL */ + 0x20002000, /* PA_SC_CLIPRECT_0_BR */ + 0x00000000, + 0x20002000, + 0x00000000, + 0x20002000, + 0x00000000, + 0x20002000, + 0xaaaaaaaa, /* PA_SC_EDGERULE */ + 0x00000000, /* PA_SU_HARDWARE_SCREEN_OFFSET */ + 0x0000000f, /* CB_TARGET_MASK */ + 0x0000000f, /* CB_SHADER_MASK */ + + 0xc0226900, + 0x00000094, + 0x80000000, /* PA_SC_VPORT_SCISSOR_0_TL */ + 0x20002000, /* PA_SC_VPORT_SCISSOR_0_BR */ + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x80000000, + 0x20002000, + 0x00000000, /* PA_SC_VPORT_ZMIN_0 */ + 0x3f800000, /* PA_SC_VPORT_ZMAX_0 */ + + 0xc0016900, + 0x000000d4, + 0x00000000, /* SX_MISC */ + + 0xc0026900, + 0x000000d9, + 0x00000000, /* CP_RINGID */ + 0x00000000, /* CP_VMID */ + + 0xc0096900, + 0x00000100, + 0x00ffffff, /* VGT_MAX_VTX_INDX */ + 0x00000000, /* VGT_MIN_VTX_INDX */ + 0x00000000, /* VGT_INDX_OFFSET */ + 0x00000000, /* VGT_MULTI_PRIM_IB_RESET_INDX */ + 0x00000000, /* SX_ALPHA_TEST_CONTROL */ + 0x00000000, /* CB_BLEND_RED */ + 0x00000000, /* CB_BLEND_GREEN */ + 0x00000000, /* CB_BLEND_BLUE */ + 0x00000000, /* CB_BLEND_ALPHA */ + + 0xc0016900, + 0x00000187, + 0x00000100, /* SPI_VS_OUT_ID_0 */ + + 0xc0026900, + 0x00000191, + 0x00000100, /* SPI_PS_INPUT_CNTL_0 */ + 0x00000101, /* SPI_PS_INPUT_CNTL_1 */ + + 0xc0016900, + 0x000001b1, + 0x00000000, /* SPI_VS_OUT_CONFIG */ + + 0xc0106900, + 0x000001b3, + 0x20000001, /* SPI_PS_IN_CONTROL_0 */ + 0x00000000, /* SPI_PS_IN_CONTROL_1 */ + 0x00000000, /* SPI_INTERP_CONTROL_0 */ + 0x00000000, /* SPI_INPUT_Z */ + 0x00000000, /* SPI_FOG_CNTL */ + 0x00100000, /* SPI_BARYC_CNTL */ + 0x00000000, /* SPI_PS_IN_CONTROL_2 */ + 0x00000000, /* SPI_COMPUTE_INPUT_CNTL */ + 0x00000000, /* SPI_COMPUTE_NUM_THREAD_X */ + 0x00000000, /* SPI_COMPUTE_NUM_THREAD_Y */ + 0x00000000, /* SPI_COMPUTE_NUM_THREAD_Z */ + 0x00000000, /* SPI_GPR_MGMT */ + 0x00000000, /* SPI_LDS_MGMT */ + 0x00000000, /* SPI_STACK_MGMT */ + 0x00000000, /* SPI_WAVE_MGMT_1 */ + 0x00000000, /* SPI_WAVE_MGMT_2 */ + + 0xc0016900, + 0x000001e0, + 0x00000000, /* CB_BLEND0_CONTROL */ + + 0xc00e6900, + 0x00000200, + 0x00000000, /* DB_DEPTH_CONTROL */ + 0x00000000, /* DB_EQAA */ + 0x00cc0010, /* CB_COLOR_CONTROL */ + 0x00000210, /* DB_SHADER_CONTROL */ + 0x00010000, /* PA_CL_CLIP_CNTL */ + 0x00000004, /* PA_SU_SC_MODE_CNTL */ + 0x00000100, /* PA_CL_VTE_CNTL */ + 0x00000000, /* PA_CL_VS_OUT_CNTL */ + 0x00000000, /* PA_CL_NANINF_CNTL */ + 0x00000000, /* PA_SU_LINE_STIPPLE_CNTL */ + 0x00000000, /* PA_SU_LINE_STIPPLE_SCALE */ + 0x00000000, /* PA_SU_PRIM_FILTER_CNTL */ + 0x00000000, /* */ + 0x00000000, /* */ + + 0xc0026900, + 0x00000229, + 0x00000000, /* SQ_PGM_START_FS */ + 0x00000000, + + 0xc0016900, + 0x0000023b, + 0x00000000, /* SQ_LDS_ALLOC_PS */ + + 0xc0066900, + 0x00000240, + 0x00000000, /* SQ_ESGS_RING_ITEMSIZE */ + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + + 0xc0046900, + 0x00000247, + 0x00000000, /* SQ_GS_VERT_ITEMSIZE */ + 0x00000000, + 0x00000000, + 0x00000000, + + 0xc0116900, + 0x00000280, + 0x00000000, /* PA_SU_POINT_SIZE */ + 0x00000000, /* PA_SU_POINT_MINMAX */ + 0x00000008, /* PA_SU_LINE_CNTL */ + 0x00000000, /* PA_SC_LINE_STIPPLE */ + 0x00000000, /* VGT_OUTPUT_PATH_CNTL */ + 0x00000000, /* VGT_HOS_CNTL */ + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, /* VGT_GS_MODE */ + + 0xc0026900, + 0x00000292, + 0x00000000, /* PA_SC_MODE_CNTL_0 */ + 0x00000000, /* PA_SC_MODE_CNTL_1 */ + + 0xc0016900, + 0x000002a1, + 0x00000000, /* VGT_PRIMITIVEID_EN */ + + 0xc0016900, + 0x000002a5, + 0x00000000, /* VGT_MULTI_PRIM_IB_RESET_EN */ + + 0xc0026900, + 0x000002a8, + 0x00000000, /* VGT_INSTANCE_STEP_RATE_0 */ + 0x00000000, + + 0xc0026900, + 0x000002ad, + 0x00000000, /* VGT_REUSE_OFF */ + 0x00000000, + + 0xc0016900, + 0x000002d5, + 0x00000000, /* VGT_SHADER_STAGES_EN */ + + 0xc0016900, + 0x000002dc, + 0x0000aa00, /* DB_ALPHA_TO_MASK */ + + 0xc0066900, + 0x000002de, + 0x00000000, /* PA_SU_POLY_OFFSET_DB_FMT_CNTL */ + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + + 0xc0026900, + 0x000002e5, + 0x00000000, /* VGT_STRMOUT_CONFIG */ + 0x00000000, + + 0xc01b6900, + 0x000002f5, + 0x76543210, /* PA_SC_CENTROID_PRIORITY_0 */ + 0xfedcba98, /* PA_SC_CENTROID_PRIORITY_1 */ + 0x00000000, /* PA_SC_LINE_CNTL */ + 0x00000000, /* PA_SC_AA_CONFIG */ + 0x00000005, /* PA_SU_VTX_CNTL */ + 0x3f800000, /* PA_CL_GB_VERT_CLIP_ADJ */ + 0x3f800000, /* PA_CL_GB_VERT_DISC_ADJ */ + 0x3f800000, /* PA_CL_GB_HORZ_CLIP_ADJ */ + 0x3f800000, /* PA_CL_GB_HORZ_DISC_ADJ */ + 0x00000000, /* PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_0 */ + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0x00000000, + 0xffffffff, /* PA_SC_AA_MASK_X0Y0_X1Y0 */ + 0xffffffff, + + 0xc0026900, + 0x00000316, + 0x0000000e, /* VGT_VERTEX_REUSE_BLOCK_CNTL */ + 0x00000010, /* */ +}; -extern const u32 cayman_ps_size, cayman_vs_size; -extern const u32 cayman_default_size; +static const u32 cayman_default_size = ARRAY_SIZE(cayman_default_state); #endif -- cgit v1.2.3 From 6f90a49bc0c9a2d8590dea3323c9b0c61ee78add Mon Sep 17 00:00:00 2001 From: Grigory Vasilyev Date: Fri, 8 Apr 2022 10:58:17 +0300 Subject: drm/amdgpu: Fix incorrect enum type Instead of the 'amdgpu_ring_priority_level' type, the 'amdgpu_gfx_pipe_priority' type was used, which is an error when setting ring priority. This is a minor error, but may cause problems in the future. Instead of AMDGPU_RING_PRIO_2 = 2, we can use AMDGPU_RING_PRIO_MAX = 3, but AMDGPU_RING_PRIO_2 = 2 is used for compatibility with AMDGPU_GFX_PIPE_PRIO_HIGH = 2, and not change the behavior of the code. Signed-off-by: Grigory Vasilyev Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c | 2 +- drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index 5147358f2400..54446162db8b 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -4741,7 +4741,7 @@ static int gfx_v10_0_compute_ring_init(struct amdgpu_device *adev, int ring_id, + ((ring->me - 1) * adev->gfx.mec.num_pipe_per_mec) + ring->pipe; hw_prio = amdgpu_gfx_is_high_priority_compute_queue(adev, ring) ? - AMDGPU_GFX_PIPE_PRIO_HIGH : AMDGPU_GFX_PIPE_PRIO_NORMAL; + AMDGPU_RING_PRIO_2 : AMDGPU_RING_PRIO_DEFAULT; /* type-2 packets are deprecated on MEC, use type-3 instead */ r = amdgpu_ring_init(adev, ring, 1024, &adev->gfx.eop_irq, irq_type, hw_prio, NULL); diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 46d436be74eb..25dc729d0ec2 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -1925,7 +1925,7 @@ static int gfx_v8_0_compute_ring_init(struct amdgpu_device *adev, int ring_id, + ring->pipe; hw_prio = amdgpu_gfx_is_high_priority_compute_queue(adev, ring) ? - AMDGPU_GFX_PIPE_PRIO_HIGH : AMDGPU_RING_PRIO_DEFAULT; + AMDGPU_RING_PRIO_2 : AMDGPU_RING_PRIO_DEFAULT; /* type-2 packets are deprecated on MEC, use type-3 instead */ r = amdgpu_ring_init(adev, ring, 1024, &adev->gfx.eop_irq, irq_type, hw_prio, NULL); diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 26535ebec498..d58fd83524ac 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -2276,7 +2276,7 @@ static int gfx_v9_0_compute_ring_init(struct amdgpu_device *adev, int ring_id, + ((ring->me - 1) * adev->gfx.mec.num_pipe_per_mec) + ring->pipe; hw_prio = amdgpu_gfx_is_high_priority_compute_queue(adev, ring) ? - AMDGPU_GFX_PIPE_PRIO_HIGH : AMDGPU_GFX_PIPE_PRIO_NORMAL; + AMDGPU_RING_PRIO_2 : AMDGPU_RING_PRIO_DEFAULT; /* type-2 packets are deprecated on MEC, use type-3 instead */ return amdgpu_ring_init(adev, ring, 1024, &adev->gfx.eop_irq, irq_type, hw_prio, NULL); -- cgit v1.2.3 From 1754cea1763e2bdc6a2153220440fe9aa9e0f2c9 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Thu, 7 Apr 2022 15:46:06 -0400 Subject: drm/amd/display: fix 64 bit divide in freesync code Use div_u64() rather than a a 64 bit divide. Fixes: 3fe5739db48843 ("drm/amd/display: Add flip interval workaround") Reviewed-by: Nathan Chancellor Reported-by: kernel test robot Signed-off-by: Alex Deucher Cc: Angus Wang Cc: Anthony Koo Cc: Aric Cyr Cc: Nathan Chancellor --- drivers/gpu/drm/amd/display/modules/freesync/freesync.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c index 0130f1879116..d2d76ce56f89 100644 --- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c +++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c @@ -1239,7 +1239,7 @@ void mod_freesync_handle_v_update(struct mod_freesync *mod_freesync, if (in_out_vrr->supported == false) return; - cur_timestamp_in_us = dm_get_timestamp(core_freesync->dc->ctx)/10; + cur_timestamp_in_us = div_u64(dm_get_timestamp(core_freesync->dc->ctx), 10); in_out_vrr->flip_interval.vsyncs_between_flip++; in_out_vrr->flip_interval.v_update_timestamp_in_us = cur_timestamp_in_us; -- cgit v1.2.3 From 05eee31c0821c7209409ef0991b95df3511c0326 Mon Sep 17 00:00:00 2001 From: "Stanley.Yang" Date: Thu, 7 Apr 2022 23:56:07 +0800 Subject: drm/amdgpu: add umc query error status function In order to debug ras error, driver will print IPID/SYND/MISC0 register value if detect correctable or uncorrectable error. Provide umc_query_error_status_helper function to reduce code redundancy. Signed-off-by: Stanley.Yang Reviewed-by: Tao Zhou Reviewed-by: Hawking Zhang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/umc_v6_7.c | 106 +++++++++++++++------------------- 1 file changed, 48 insertions(+), 58 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c b/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c index c45d9c14ecbc..606892dbea1c 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c +++ b/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c @@ -64,21 +64,62 @@ static inline uint32_t get_umc_v6_7_channel_index(struct amdgpu_device *adev, return adev->umc.channel_idx_tbl[umc_inst * adev->umc.channel_inst_num + ch_inst]; } +static void umc_v6_7_query_error_status_helper(struct amdgpu_device *adev, + uint64_t mc_umc_status, uint32_t umc_reg_offset) +{ + uint32_t mc_umc_addr; + uint64_t reg_value; + + if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Deferred) == 1) + dev_info(adev->dev, "Deferred error, no user action is needed.\n"); + + if (mc_umc_status) + dev_info(adev->dev, "MCA STATUS 0x%llx, umc_reg_offset 0x%x\n", mc_umc_status, umc_reg_offset); + + /* print IPID registers value */ + mc_umc_addr = + SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_IPIDT0); + reg_value = RREG64_PCIE((mc_umc_addr + umc_reg_offset) * 4); + if (reg_value) + dev_info(adev->dev, "MCA IPID 0x%llx, umc_reg_offset 0x%x\n", reg_value, umc_reg_offset); + + /* print SYND registers value */ + mc_umc_addr = + SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_SYNDT0); + reg_value = RREG64_PCIE((mc_umc_addr + umc_reg_offset) * 4); + if (reg_value) + dev_info(adev->dev, "MCA SYND 0x%llx, umc_reg_offset 0x%x\n", reg_value, umc_reg_offset); + + /* print MISC0 registers value */ + mc_umc_addr = + SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_MISC0T0); + reg_value = RREG64_PCIE((mc_umc_addr + umc_reg_offset) * 4); + if (reg_value) + dev_info(adev->dev, "MCA MISC0 0x%llx, umc_reg_offset 0x%x\n", reg_value, umc_reg_offset); +} + static void umc_v6_7_ecc_info_query_correctable_error_count(struct amdgpu_device *adev, uint32_t umc_inst, uint32_t ch_inst, unsigned long *error_count) { uint64_t mc_umc_status; uint32_t eccinfo_table_idx; + uint32_t umc_reg_offset; struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + umc_reg_offset = get_umc_v6_7_reg_offset(adev, + umc_inst, ch_inst); + eccinfo_table_idx = umc_inst * adev->umc.channel_inst_num + ch_inst; /* check for SRAM correctable error MCUMC_STATUS is a 64 bit register */ mc_umc_status = ras->umc_ecc.ecc[eccinfo_table_idx].mca_umc_status; if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1 && - REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1) + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1) { *error_count += 1; + + umc_v6_7_query_error_status_helper(adev, mc_umc_status, umc_reg_offset); + } } static void umc_v6_7_ecc_info_querry_uncorrectable_error_count(struct amdgpu_device *adev, @@ -88,8 +129,6 @@ static void umc_v6_7_ecc_info_querry_uncorrectable_error_count(struct amdgpu_dev uint64_t mc_umc_status; uint32_t eccinfo_table_idx; uint32_t umc_reg_offset; - uint32_t mc_umc_addr; - uint64_t reg_value; struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); umc_reg_offset = get_umc_v6_7_reg_offset(adev, @@ -106,32 +145,7 @@ static void umc_v6_7_ecc_info_querry_uncorrectable_error_count(struct amdgpu_dev REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, TCC) == 1)) { *error_count += 1; - if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Deferred) == 1) - dev_info(adev->dev, "Deferred error, no user action is needed.\n"); - - if (mc_umc_status) - dev_info(adev->dev, "MCA STATUS 0x%llx, umc_reg_offset 0x%x\n", mc_umc_status, umc_reg_offset); - - /* print IPID registers value */ - mc_umc_addr = - SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_IPIDT0); - reg_value = RREG64_PCIE((mc_umc_addr + umc_reg_offset) * 4); - if (reg_value) - dev_info(adev->dev, "MCA IPID 0x%llx, umc_reg_offset 0x%x\n", reg_value, umc_reg_offset); - - /* print SYND registers value */ - mc_umc_addr = - SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_SYNDT0); - reg_value = RREG64_PCIE((mc_umc_addr + umc_reg_offset) * 4); - if (reg_value) - dev_info(adev->dev, "MCA SYND 0x%llx, umc_reg_offset 0x%x\n", reg_value, umc_reg_offset); - - /* print MISC0 registers value */ - mc_umc_addr = - SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_MISC0T0); - reg_value = RREG64_PCIE((mc_umc_addr + umc_reg_offset) * 4); - if (reg_value) - dev_info(adev->dev, "MCA MISC0 0x%llx, umc_reg_offset 0x%x\n", reg_value, umc_reg_offset); + umc_v6_7_query_error_status_helper(adev, mc_umc_status, umc_reg_offset); } } @@ -277,8 +291,11 @@ static void umc_v6_7_query_correctable_error_count(struct amdgpu_device *adev, MCUMC_STATUS is a 64 bit register */ mc_umc_status = RREG64_PCIE((mc_umc_status_addr + umc_reg_offset) * 4); if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1 && - REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1) + REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1) { *error_count += 1; + + umc_v6_7_query_error_status_helper(adev, mc_umc_status, umc_reg_offset); + } } static void umc_v6_7_querry_uncorrectable_error_count(struct amdgpu_device *adev, @@ -287,8 +304,6 @@ static void umc_v6_7_querry_uncorrectable_error_count(struct amdgpu_device *adev { uint64_t mc_umc_status; uint32_t mc_umc_status_addr; - uint32_t mc_umc_addr; - uint64_t reg_value; mc_umc_status_addr = SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_STATUST0); @@ -303,32 +318,7 @@ static void umc_v6_7_querry_uncorrectable_error_count(struct amdgpu_device *adev REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, TCC) == 1)) { *error_count += 1; - if (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Deferred) == 1) - dev_info(adev->dev, "Deferred error, no user action is needed.\n"); - - if (mc_umc_status) - dev_info(adev->dev, "MCA STATUS 0x%llx, umc_reg_offset 0x%x\n", mc_umc_status, umc_reg_offset); - - /* print IPID registers value */ - mc_umc_addr = - SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_IPIDT0); - reg_value = RREG64_PCIE((mc_umc_addr + umc_reg_offset) * 4); - if (reg_value) - dev_info(adev->dev, "MCA IPID 0x%llx, umc_reg_offset 0x%x\n", reg_value, umc_reg_offset); - - /* print SYND registers value */ - mc_umc_addr = - SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_SYNDT0); - reg_value = RREG64_PCIE((mc_umc_addr + umc_reg_offset) * 4); - if (reg_value) - dev_info(adev->dev, "MCA SYND 0x%llx, umc_reg_offset 0x%x\n", reg_value, umc_reg_offset); - - /* print MISC0 registers value */ - mc_umc_addr = - SOC15_REG_OFFSET(UMC, 0, regMCA_UMC_UMC0_MCUMC_MISC0T0); - reg_value = RREG64_PCIE((mc_umc_addr + umc_reg_offset) * 4); - if (reg_value) - dev_info(adev->dev, "MCA MISC0 0x%llx, umc_reg_offset 0x%x\n", reg_value, umc_reg_offset); + umc_v6_7_query_error_status_helper(adev, mc_umc_status, umc_reg_offset); } } -- cgit v1.2.3 From a0bd69e1268ba74c712e91afd6afb283f0b4d7ae Mon Sep 17 00:00:00 2001 From: Eric Yang Date: Sat, 19 Mar 2022 16:34:24 -0400 Subject: drm/amd/display: undo clearing of z10 related function pointers [Why] Z10 and S0i3 have some shared path. Previous code clean up , incorrectly removed these pointers, which breaks s0i3 restore [How] Do not clear the function pointers based on Z10 disable. Reviewed-by: Nicholas Kazlauskas Acked-by: Pavle Kotarac Signed-off-by: Eric Yang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn31/dcn31_init.c | 5 ----- 1 file changed, 5 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_init.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_init.c index d7559e5a99ce..e708f07fe75a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_init.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_init.c @@ -153,9 +153,4 @@ void dcn31_hw_sequencer_construct(struct dc *dc) dc->hwss.init_hw = dcn20_fpga_init_hw; dc->hwseq->funcs.init_pipes = NULL; } - if (dc->debug.disable_z10) { - /*hw not support z10 or sw disable it*/ - dc->hwss.z10_restore = NULL; - dc->hwss.z10_save_init = NULL; - } } -- cgit v1.2.3 From 2513ed4f937999c0446fd824f7564f76b697d722 Mon Sep 17 00:00:00 2001 From: Josip Pavic Date: Thu, 24 Mar 2022 12:08:43 -0400 Subject: drm/amd/display: do not wait for mpc idle if tg is disabled [Why] When booting, the driver waits for the MPC idle bit to be set as part of pipe initialization. However, on some systems this occurs before OTG is enabled, and since the MPC idle bit won't be set until the vupdate signal occurs (which requires OTG to be enabled), this never happens and the wait times out. This can add hundreds of milliseconds to the boot time. [How] Do not wait for mpc idle if tg is disabled Reviewed-by: Jun Lei Acked-by: Pavle Kotarac Signed-off-by: Josip Pavic Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index 724c04fe004b..c84011caa73c 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c @@ -3189,7 +3189,8 @@ void dcn10_wait_for_mpcc_disconnect( if (pipe_ctx->stream_res.opp->mpcc_disconnect_pending[mpcc_inst]) { struct hubp *hubp = get_hubp_by_inst(res_pool, mpcc_inst); - res_pool->mpc->funcs->wait_for_idle(res_pool->mpc, mpcc_inst); + if (pipe_ctx->stream_res.tg->funcs->is_tg_enabled(pipe_ctx->stream_res.tg)) + res_pool->mpc->funcs->wait_for_idle(res_pool->mpc, mpcc_inst); pipe_ctx->stream_res.opp->mpcc_disconnect_pending[mpcc_inst] = false; hubp->funcs->set_blank(hubp, true); } -- cgit v1.2.3 From 5d5af34072c8b11f60960c3bea57ff9de5877791 Mon Sep 17 00:00:00 2001 From: Saaem Rizvi Date: Mon, 28 Mar 2022 14:25:16 -0400 Subject: drm/amd/display: Disabling Z10 on DCN31 [WHY] Z10 is should not be enabled by default on DCN31. [HOW] Using DC debug flags to disable Z10 by default on DCN31. Reviewed-by: Eric Yang Acked-by: Pavle Kotarac Signed-off-by: Saaem Rizvi Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c | 1 + 1 file changed, 1 insertion(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c index f27262417abe..6cc580be7c79 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c @@ -885,6 +885,7 @@ static const struct dc_debug_options debug_defaults_drv = { .afmt = true, } }, + .disable_z10 = true, .optimize_edp_link_rate = true, .enable_sw_cntl_psr = true, .apply_vendor_specific_lttpr_wa = true, -- cgit v1.2.3 From 1b5c30bf41586fdf4f2313d0b37545975568e29f Mon Sep 17 00:00:00 2001 From: Max Erenberg Date: Tue, 23 Nov 2021 15:10:16 -0500 Subject: drm/amd/display: Disallow entering PSR when panel is disconnected [WHY] The dGPU cannot enter PSR when it is not connected to a panel. [HOW] Added a check to dc_link_set_psr_allow_active which returns early if panel is disconnected. Reviewed-by: Harry Vanzylldejong Reviewed-by: Evgenii Krasnikov Reviewed-by: Nicholas Choi Acked-by: Pavle Kotarac Signed-off-by: Max Erenberg Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc_link.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index 47b67fd1e84c..22f2d88fab99 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -3079,6 +3079,11 @@ bool dc_link_set_psr_allow_active(struct dc_link *link, const bool *allow_active if (!dc_get_edp_link_panel_inst(dc, link, &panel_inst)) return false; + if (allow_active && link->type == dc_connection_none) { + // Don't enter PSR if panel is not connected + return false; + } + /* Set power optimization flag */ if (power_opts && link->psr_settings.psr_power_opt != *power_opts) { link->psr_settings.psr_power_opt = *power_opts; -- cgit v1.2.3 From 6866a60a8ef7f320eeaf737d4e2428ea40d252ff Mon Sep 17 00:00:00 2001 From: Tom Rix Date: Sat, 9 Apr 2022 13:11:31 -0400 Subject: drm/radeon: remove r600_blit_shaders.[c|h] The only use of the global variables in r600_blit_shaders.c were in the old drivers/gpu/drm/radeon/r600_blit.c This file was removed in commit 8333f607a631 ("drm/radeon: remove UMS support") So remove the r600_blit_shaders.[c|h] files Signed-off-by: Tom Rix Signed-off-by: Alex Deucher --- drivers/gpu/drm/radeon/Makefile | 2 +- drivers/gpu/drm/radeon/r600_blit_shaders.c | 719 ----------------------------- drivers/gpu/drm/radeon/r600_blit_shaders.h | 38 -- 3 files changed, 1 insertion(+), 758 deletions(-) delete mode 100644 drivers/gpu/drm/radeon/r600_blit_shaders.c delete mode 100644 drivers/gpu/drm/radeon/r600_blit_shaders.h (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile index 2425a3612d6c..4deedaacd655 100644 --- a/drivers/gpu/drm/radeon/Makefile +++ b/drivers/gpu/drm/radeon/Makefile @@ -39,7 +39,7 @@ radeon-y += radeon_device.o radeon_asic.o radeon_kms.o \ radeon_clocks.o radeon_fb.o radeon_gem.o radeon_ring.o radeon_irq_kms.o \ radeon_cs.o radeon_bios.o radeon_benchmark.o r100.o r300.o r420.o \ rs400.o rs600.o rs690.o rv515.o r520.o r600.o rv770.o radeon_test.o \ - r200.o radeon_legacy_tv.o r600_cs.o r600_blit_shaders.o \ + r200.o radeon_legacy_tv.o r600_cs.o \ radeon_pm.o atombios_dp.o r600_hdmi.o dce3_1_afmt.o \ evergreen.o evergreen_cs.o evergreen_blit_shaders.o \ evergreen_hdmi.o radeon_trace_points.o ni.o \ diff --git a/drivers/gpu/drm/radeon/r600_blit_shaders.c b/drivers/gpu/drm/radeon/r600_blit_shaders.c deleted file mode 100644 index 443cbe59b274..000000000000 --- a/drivers/gpu/drm/radeon/r600_blit_shaders.c +++ /dev/null @@ -1,719 +0,0 @@ -/* - * Copyright 2009 Advanced Micro Devices, Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice (including the next - * paragraph) shall be included in all copies or substantial portions of the - * Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER - * DEALINGS IN THE SOFTWARE. - * - * Authors: - * Alex Deucher - */ - -#include -#include -#include - -/* - * R6xx+ cards need to use the 3D engine to blit data which requires - * quite a bit of hw state setup. Rather than pull the whole 3D driver - * (which normally generates the 3D state) into the DRM, we opt to use - * statically generated state tables. The register state and shaders - * were hand generated to support blitting functionality. See the 3D - * driver or documentation for descriptions of the registers and - * shader instructions. - */ - -const u32 r6xx_default_state[] = -{ - 0xc0002400, /* START_3D_CMDBUF */ - 0x00000000, - - 0xc0012800, /* CONTEXT_CONTROL */ - 0x80000000, - 0x80000000, - - 0xc0016800, - 0x00000010, - 0x00008000, /* WAIT_UNTIL */ - - 0xc0016800, - 0x00000542, - 0x07000003, /* TA_CNTL_AUX */ - - 0xc0016800, - 0x000005c5, - 0x00000000, /* VC_ENHANCE */ - - 0xc0016800, - 0x00000363, - 0x00000000, /* SQ_DYN_GPR_CNTL_PS_FLUSH_REQ */ - - 0xc0016800, - 0x0000060c, - 0x82000000, /* DB_DEBUG */ - - 0xc0016800, - 0x0000060e, - 0x01020204, /* DB_WATERMARKS */ - - 0xc0026f00, - 0x00000000, - 0x00000000, /* SQ_VTX_BASE_VTX_LOC */ - 0x00000000, /* SQ_VTX_START_INST_LOC */ - - 0xc0096900, - 0x0000022a, - 0x00000000, /* SQ_ESGS_RING_ITEMSIZE */ - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - - 0xc0016900, - 0x00000004, - 0x00000000, /* DB_DEPTH_INFO */ - - 0xc0026900, - 0x0000000a, - 0x00000000, /* DB_STENCIL_CLEAR */ - 0x00000000, /* DB_DEPTH_CLEAR */ - - 0xc0016900, - 0x00000200, - 0x00000000, /* DB_DEPTH_CONTROL */ - - 0xc0026900, - 0x00000343, - 0x00000060, /* DB_RENDER_CONTROL */ - 0x00000040, /* DB_RENDER_OVERRIDE */ - - 0xc0016900, - 0x00000351, - 0x0000aa00, /* DB_ALPHA_TO_MASK */ - - 0xc00f6900, - 0x00000100, - 0x00000800, /* VGT_MAX_VTX_INDX */ - 0x00000000, /* VGT_MIN_VTX_INDX */ - 0x00000000, /* VGT_INDX_OFFSET */ - 0x00000000, /* VGT_MULTI_PRIM_IB_RESET_INDX */ - 0x00000000, /* SX_ALPHA_TEST_CONTROL */ - 0x00000000, /* CB_BLEND_RED */ - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, /* CB_FOG_RED */ - 0x00000000, - 0x00000000, - 0x00000000, /* DB_STENCILREFMASK */ - 0x00000000, /* DB_STENCILREFMASK_BF */ - 0x00000000, /* SX_ALPHA_REF */ - - 0xc0046900, - 0x0000030c, - 0x01000000, /* CB_CLRCMP_CNTL */ - 0x00000000, - 0x00000000, - 0x00000000, - - 0xc0046900, - 0x00000048, - 0x3f800000, /* CB_CLEAR_RED */ - 0x00000000, - 0x3f800000, - 0x3f800000, - - 0xc0016900, - 0x00000080, - 0x00000000, /* PA_SC_WINDOW_OFFSET */ - - 0xc00a6900, - 0x00000083, - 0x0000ffff, /* PA_SC_CLIP_RECT_RULE */ - 0x00000000, /* PA_SC_CLIPRECT_0_TL */ - 0x20002000, - 0x00000000, - 0x20002000, - 0x00000000, - 0x20002000, - 0x00000000, - 0x20002000, - 0x00000000, /* PA_SC_EDGERULE */ - - 0xc0406900, - 0x00000094, - 0x80000000, /* PA_SC_VPORT_SCISSOR_0_TL */ - 0x20002000, /* PA_SC_VPORT_SCISSOR_0_BR */ - 0x80000000, /* PA_SC_VPORT_SCISSOR_1_TL */ - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x00000000, /* PA_SC_VPORT_ZMIN_0 */ - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - - 0xc0026900, - 0x00000292, - 0x00000000, /* PA_SC_MPASS_PS_CNTL */ - 0x00004010, /* PA_SC_MODE_CNTL */ - - 0xc0096900, - 0x00000300, - 0x00000000, /* PA_SC_LINE_CNTL */ - 0x00000000, /* PA_SC_AA_CONFIG */ - 0x0000002d, /* PA_SU_VTX_CNTL */ - 0x3f800000, /* PA_CL_GB_VERT_CLIP_ADJ */ - 0x3f800000, - 0x3f800000, - 0x3f800000, - 0x00000000, /* PA_SC_SAMPLE_LOCS_MCTX */ - 0x00000000, - - 0xc0016900, - 0x00000312, - 0xffffffff, /* PA_SC_AA_MASK */ - - 0xc0066900, - 0x0000037e, - 0x00000000, /* PA_SU_POLY_OFFSET_DB_FMT_CNTL */ - 0x00000000, /* PA_SU_POLY_OFFSET_CLAMP */ - 0x00000000, /* PA_SU_POLY_OFFSET_FRONT_SCALE */ - 0x00000000, /* PA_SU_POLY_OFFSET_FRONT_OFFSET */ - 0x00000000, /* PA_SU_POLY_OFFSET_BACK_SCALE */ - 0x00000000, /* PA_SU_POLY_OFFSET_BACK_OFFSET */ - - 0xc0046900, - 0x000001b6, - 0x00000000, /* SPI_INPUT_Z */ - 0x00000000, /* SPI_FOG_CNTL */ - 0x00000000, /* SPI_FOG_FUNC_SCALE */ - 0x00000000, /* SPI_FOG_FUNC_BIAS */ - - 0xc0016900, - 0x00000225, - 0x00000000, /* SQ_PGM_START_FS */ - - 0xc0016900, - 0x00000229, - 0x00000000, /* SQ_PGM_RESOURCES_FS */ - - 0xc0016900, - 0x00000237, - 0x00000000, /* SQ_PGM_CF_OFFSET_FS */ - - 0xc0026900, - 0x000002a8, - 0x00000000, /* VGT_INSTANCE_STEP_RATE_0 */ - 0x00000000, /* VGT_INSTANCE_STEP_RATE_1 */ - - 0xc0116900, - 0x00000280, - 0x00000000, /* PA_SU_POINT_SIZE */ - 0x00000000, /* PA_SU_POINT_MINMAX */ - 0x00000008, /* PA_SU_LINE_CNTL */ - 0x00000000, /* PA_SC_LINE_STIPPLE */ - 0x00000000, /* VGT_OUTPUT_PATH_CNTL */ - 0x00000000, /* VGT_HOS_CNTL */ - 0x00000000, /* VGT_HOS_MAX_TESS_LEVEL */ - 0x00000000, /* VGT_HOS_MIN_TESS_LEVEL */ - 0x00000000, /* VGT_HOS_REUSE_DEPTH */ - 0x00000000, /* VGT_GROUP_PRIM_TYPE */ - 0x00000000, /* VGT_GROUP_FIRST_DECR */ - 0x00000000, /* VGT_GROUP_DECR */ - 0x00000000, /* VGT_GROUP_VECT_0_CNTL */ - 0x00000000, /* VGT_GROUP_VECT_1_CNTL */ - 0x00000000, /* VGT_GROUP_VECT_0_FMT_CNTL */ - 0x00000000, /* VGT_GROUP_VECT_1_FMT_CNTL */ - 0x00000000, /* VGT_GS_MODE */ - - 0xc0016900, - 0x000002a1, - 0x00000000, /* VGT_PRIMITIVEID_EN */ - - 0xc0016900, - 0x000002a5, - 0x00000000, /* VGT_MULTI_PRIM_ID_RESET_EN */ - - 0xc0036900, - 0x000002ac, - 0x00000000, /* VGT_STRMOUT_EN */ - 0x00000000, /* VGT_REUSE_OFF */ - 0x00000000, /* VGT_VTX_CNT_EN */ - - 0xc0016900, - 0x000000d4, - 0x00000000, /* SX_MISC */ - - 0xc0016900, - 0x000002c8, - 0x00000000, /* VGT_STRMOUT_BUFFER_EN */ - - 0xc0076900, - 0x00000202, - 0x00cc0000, /* CB_COLOR_CONTROL */ - 0x00000210, /* DB_SHADER_CNTL */ - 0x00010000, /* PA_CL_CLIP_CNTL */ - 0x00000244, /* PA_SU_SC_MODE_CNTL */ - 0x00000100, /* PA_CL_VTE_CNTL */ - 0x00000000, /* PA_CL_VS_OUT_CNTL */ - 0x00000000, /* PA_CL_NANINF_CNTL */ - - 0xc0026900, - 0x0000008e, - 0x0000000f, /* CB_TARGET_MASK */ - 0x0000000f, /* CB_SHADER_MASK */ - - 0xc0016900, - 0x000001e8, - 0x00000001, /* CB_SHADER_CONTROL */ - - 0xc0016900, - 0x00000185, - 0x00000000, /* SPI_VS_OUT_ID_0 */ - - 0xc0016900, - 0x00000191, - 0x00000b00, /* SPI_PS_INPUT_CNTL_0 */ - - 0xc0056900, - 0x000001b1, - 0x00000000, /* SPI_VS_OUT_CONFIG */ - 0x00000000, /* SPI_THREAD_GROUPING */ - 0x00000001, /* SPI_PS_IN_CONTROL_0 */ - 0x00000000, /* SPI_PS_IN_CONTROL_1 */ - 0x00000000, /* SPI_INTERP_CONTROL_0 */ - - 0xc0036e00, /* SET_SAMPLER */ - 0x00000000, - 0x00000012, - 0x00000000, - 0x00000000, -}; - -const u32 r7xx_default_state[] = -{ - 0xc0012800, /* CONTEXT_CONTROL */ - 0x80000000, - 0x80000000, - - 0xc0016800, - 0x00000010, - 0x00008000, /* WAIT_UNTIL */ - - 0xc0016800, - 0x00000542, - 0x07000002, /* TA_CNTL_AUX */ - - 0xc0016800, - 0x000005c5, - 0x00000000, /* VC_ENHANCE */ - - 0xc0016800, - 0x00000363, - 0x00004000, /* SQ_DYN_GPR_CNTL_PS_FLUSH_REQ */ - - 0xc0016800, - 0x0000060c, - 0x00000000, /* DB_DEBUG */ - - 0xc0016800, - 0x0000060e, - 0x00420204, /* DB_WATERMARKS */ - - 0xc0026f00, - 0x00000000, - 0x00000000, /* SQ_VTX_BASE_VTX_LOC */ - 0x00000000, /* SQ_VTX_START_INST_LOC */ - - 0xc0096900, - 0x0000022a, - 0x00000000, /* SQ_ESGS_RING_ITEMSIZE */ - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - 0x00000000, - - 0xc0016900, - 0x00000004, - 0x00000000, /* DB_DEPTH_INFO */ - - 0xc0026900, - 0x0000000a, - 0x00000000, /* DB_STENCIL_CLEAR */ - 0x00000000, /* DB_DEPTH_CLEAR */ - - 0xc0016900, - 0x00000200, - 0x00000000, /* DB_DEPTH_CONTROL */ - - 0xc0026900, - 0x00000343, - 0x00000060, /* DB_RENDER_CONTROL */ - 0x00000000, /* DB_RENDER_OVERRIDE */ - - 0xc0016900, - 0x00000351, - 0x0000aa00, /* DB_ALPHA_TO_MASK */ - - 0xc0096900, - 0x00000100, - 0x00000800, /* VGT_MAX_VTX_INDX */ - 0x00000000, /* VGT_MIN_VTX_INDX */ - 0x00000000, /* VGT_INDX_OFFSET */ - 0x00000000, /* VGT_MULTI_PRIM_IB_RESET_INDX */ - 0x00000000, /* SX_ALPHA_TEST_CONTROL */ - 0x00000000, /* CB_BLEND_RED */ - 0x00000000, - 0x00000000, - 0x00000000, - - 0xc0036900, - 0x0000010c, - 0x00000000, /* DB_STENCILREFMASK */ - 0x00000000, /* DB_STENCILREFMASK_BF */ - 0x00000000, /* SX_ALPHA_REF */ - - 0xc0046900, - 0x0000030c, /* CB_CLRCMP_CNTL */ - 0x01000000, - 0x00000000, - 0x00000000, - 0x00000000, - - 0xc0016900, - 0x00000080, - 0x00000000, /* PA_SC_WINDOW_OFFSET */ - - 0xc00a6900, - 0x00000083, - 0x0000ffff, /* PA_SC_CLIP_RECT_RULE */ - 0x00000000, /* PA_SC_CLIPRECT_0_TL */ - 0x20002000, - 0x00000000, - 0x20002000, - 0x00000000, - 0x20002000, - 0x00000000, - 0x20002000, - 0xaaaaaaaa, /* PA_SC_EDGERULE */ - - 0xc0406900, - 0x00000094, - 0x80000000, /* PA_SC_VPORT_SCISSOR_0_TL */ - 0x20002000, /* PA_SC_VPORT_SCISSOR_0_BR */ - 0x80000000, /* PA_SC_VPORT_SCISSOR_1_TL */ - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x80000000, - 0x20002000, - 0x00000000, /* PA_SC_VPORT_ZMIN_0 */ - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - 0x00000000, - 0x3f800000, - - 0xc0026900, - 0x00000292, - 0x00000000, /* PA_SC_MPASS_PS_CNTL */ - 0x00514000, /* PA_SC_MODE_CNTL */ - - 0xc0096900, - 0x00000300, - 0x00000000, /* PA_SC_LINE_CNTL */ - 0x00000000, /* PA_SC_AA_CONFIG */ - 0x0000002d, /* PA_SU_VTX_CNTL */ - 0x3f800000, /* PA_CL_GB_VERT_CLIP_ADJ */ - 0x3f800000, - 0x3f800000, - 0x3f800000, - 0x00000000, /* PA_SC_SAMPLE_LOCS_MCTX */ - 0x00000000, - - 0xc0016900, - 0x00000312, - 0xffffffff, /* PA_SC_AA_MASK */ - - 0xc0066900, - 0x0000037e, - 0x00000000, /* PA_SU_POLY_OFFSET_DB_FMT_CNTL */ - 0x00000000, /* PA_SU_POLY_OFFSET_CLAMP */ - 0x00000000, /* PA_SU_POLY_OFFSET_FRONT_SCALE */ - 0x00000000, /* PA_SU_POLY_OFFSET_FRONT_OFFSET */ - 0x00000000, /* PA_SU_POLY_OFFSET_BACK_SCALE */ - 0x00000000, /* PA_SU_POLY_OFFSET_BACK_OFFSET */ - - 0xc0046900, - 0x000001b6, - 0x00000000, /* SPI_INPUT_Z */ - 0x00000000, /* SPI_FOG_CNTL */ - 0x00000000, /* SPI_FOG_FUNC_SCALE */ - 0x00000000, /* SPI_FOG_FUNC_BIAS */ - - 0xc0016900, - 0x00000225, - 0x00000000, /* SQ_PGM_START_FS */ - - 0xc0016900, - 0x00000229, - 0x00000000, /* SQ_PGM_RESOURCES_FS */ - - 0xc0016900, - 0x00000237, - 0x00000000, /* SQ_PGM_CF_OFFSET_FS */ - - 0xc0026900, - 0x000002a8, - 0x00000000, /* VGT_INSTANCE_STEP_RATE_0 */ - 0x00000000, /* VGT_INSTANCE_STEP_RATE_1 */ - - 0xc0116900, - 0x00000280, - 0x00000000, /* PA_SU_POINT_SIZE */ - 0x00000000, /* PA_SU_POINT_MINMAX */ - 0x00000008, /* PA_SU_LINE_CNTL */ - 0x00000000, /* PA_SC_LINE_STIPPLE */ - 0x00000000, /* VGT_OUTPUT_PATH_CNTL */ - 0x00000000, /* VGT_HOS_CNTL */ - 0x00000000, /* VGT_HOS_MAX_TESS_LEVEL */ - 0x00000000, /* VGT_HOS_MIN_TESS_LEVEL */ - 0x00000000, /* VGT_HOS_REUSE_DEPTH */ - 0x00000000, /* VGT_GROUP_PRIM_TYPE */ - 0x00000000, /* VGT_GROUP_FIRST_DECR */ - 0x00000000, /* VGT_GROUP_DECR */ - 0x00000000, /* VGT_GROUP_VECT_0_CNTL */ - 0x00000000, /* VGT_GROUP_VECT_1_CNTL */ - 0x00000000, /* VGT_GROUP_VECT_0_FMT_CNTL */ - 0x00000000, /* VGT_GROUP_VECT_1_FMT_CNTL */ - 0x00000000, /* VGT_GS_MODE */ - - 0xc0016900, - 0x000002a1, - 0x00000000, /* VGT_PRIMITIVEID_EN */ - - 0xc0016900, - 0x000002a5, - 0x00000000, /* VGT_MULTI_PRIM_ID_RESET_EN */ - - 0xc0036900, - 0x000002ac, - 0x00000000, /* VGT_STRMOUT_EN */ - 0x00000000, /* VGT_REUSE_OFF */ - 0x00000000, /* VGT_VTX_CNT_EN */ - - 0xc0016900, - 0x000000d4, - 0x00000000, /* SX_MISC */ - - 0xc0016900, - 0x000002c8, - 0x00000000, /* VGT_STRMOUT_BUFFER_EN */ - - 0xc0076900, - 0x00000202, - 0x00cc0000, /* CB_COLOR_CONTROL */ - 0x00000210, /* DB_SHADER_CNTL */ - 0x00010000, /* PA_CL_CLIP_CNTL */ - 0x00000244, /* PA_SU_SC_MODE_CNTL */ - 0x00000100, /* PA_CL_VTE_CNTL */ - 0x00000000, /* PA_CL_VS_OUT_CNTL */ - 0x00000000, /* PA_CL_NANINF_CNTL */ - - 0xc0026900, - 0x0000008e, - 0x0000000f, /* CB_TARGET_MASK */ - 0x0000000f, /* CB_SHADER_MASK */ - - 0xc0016900, - 0x000001e8, - 0x00000001, /* CB_SHADER_CONTROL */ - - 0xc0016900, - 0x00000185, - 0x00000000, /* SPI_VS_OUT_ID_0 */ - - 0xc0016900, - 0x00000191, - 0x00000b00, /* SPI_PS_INPUT_CNTL_0 */ - - 0xc0056900, - 0x000001b1, - 0x00000000, /* SPI_VS_OUT_CONFIG */ - 0x00000001, /* SPI_THREAD_GROUPING */ - 0x00000001, /* SPI_PS_IN_CONTROL_0 */ - 0x00000000, /* SPI_PS_IN_CONTROL_1 */ - 0x00000000, /* SPI_INTERP_CONTROL_0 */ - - 0xc0036e00, /* SET_SAMPLER */ - 0x00000000, - 0x00000012, - 0x00000000, - 0x00000000, -}; - -/* same for r6xx/r7xx */ -const u32 r6xx_vs[] = -{ - 0x00000004, - 0x81000000, - 0x0000203c, - 0x94000b08, - 0x00004000, - 0x14200b1a, - 0x00000000, - 0x00000000, - 0x3c000000, - 0x68cd1000, -#ifdef __BIG_ENDIAN - 0x000a0000, -#else - 0x00080000, -#endif - 0x00000000, -}; - -const u32 r6xx_ps[] = -{ - 0x00000002, - 0x80800000, - 0x00000000, - 0x94200688, - 0x00000010, - 0x000d1000, - 0xb0800000, - 0x00000000, -}; - -const u32 r6xx_ps_size = ARRAY_SIZE(r6xx_ps); -const u32 r6xx_vs_size = ARRAY_SIZE(r6xx_vs); -const u32 r6xx_default_size = ARRAY_SIZE(r6xx_default_state); -const u32 r7xx_default_size = ARRAY_SIZE(r7xx_default_state); diff --git a/drivers/gpu/drm/radeon/r600_blit_shaders.h b/drivers/gpu/drm/radeon/r600_blit_shaders.h deleted file mode 100644 index f437d36dd98c..000000000000 --- a/drivers/gpu/drm/radeon/r600_blit_shaders.h +++ /dev/null @@ -1,38 +0,0 @@ -/* - * Copyright 2009 Advanced Micro Devices, Inc. - * Copyright 2009 Red Hat Inc. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice (including the next - * paragraph) shall be included in all copies or substantial portions of the - * Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE COPYRIGHT HOLDER(S) AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, DAMAGES OR - * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, - * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER - * DEALINGS IN THE SOFTWARE. - * - */ - -#ifndef R600_BLIT_SHADERS_H -#define R600_BLIT_SHADERS_H - -extern const u32 r6xx_ps[]; -extern const u32 r6xx_vs[]; -extern const u32 r7xx_default_state[]; -extern const u32 r6xx_default_state[]; - - -extern const u32 r6xx_ps_size, r6xx_vs_size; -extern const u32 r6xx_default_size, r7xx_default_size; - -#endif -- cgit v1.2.3 From 9e6a04651e0231c4ad660e1580b40e56d6d5fde5 Mon Sep 17 00:00:00 2001 From: Iswara Nagulendran Date: Wed, 5 Jan 2022 17:56:25 -0500 Subject: drm/amd/display: Disable FEC if DSC not supported for EDP [WHY] Screen was seen corrupted for a few ms when switching both ways. There was also not enough bandwidth for HDR to be enabled in HG disabled mode. This was due to FEC being enabled although DSC was not supported or disabled for the EDP. [HOW] Check for EDP DSC support in DC caps or if DSC should be disabled for EDP before enabling FEC for EDP. Reviewed-by: Harry Vanzylldejong Reviewed-by: Evgenii Krasnikov Reviewed-by: Nicholas Choi Acked-by: Pavle Kotarac Signed-off-by: Iswara Nagulendran Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc_link.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index 22f2d88fab99..c1a727ff8b1d 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -4663,8 +4663,10 @@ bool dc_link_should_enable_fec(const struct dc_link *link) link->local_sink->edid_caps.panel_patch.disable_fec) force_disable = true; else if (link->connector_signal == SIGNAL_TYPE_EDP - && link->dpcd_caps.dsc_caps.dsc_basic_caps.fields. - dsc_support.DSC_SUPPORT == false) + && (link->dpcd_caps.dsc_caps.dsc_basic_caps.fields. + dsc_support.DSC_SUPPORT == false + || link->dc->debug.disable_dsc_edp + || !link->dc->caps.edp_dsc_support)) force_disable = true; return !force_disable && dc_link_is_fec_supported(link); -- cgit v1.2.3 From da8c25929b78b39438eb62b87e213c45dfad6038 Mon Sep 17 00:00:00 2001 From: Harry VanZyllDeJong Date: Tue, 11 Jan 2022 15:11:42 -0500 Subject: drm/amd/display: Fix crash on setting VRR with no display connected [HOW&WHY] VRR was getting set at the same time the timing generator would be null when there was no display connected. Added null check to the timing generator variable so it does not get referenced if it is null. Reviewed-by: Harry Vanzylldejong Reviewed-by: Evgenii Krasnikov Reviewed-by: Nicholas Choi Acked-by: Pavle Kotarac Signed-off-by: Harry VanZyllDeJong Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index c84011caa73c..6c69fd3bc77a 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c @@ -3053,12 +3053,16 @@ void dcn10_set_drr(struct pipe_ctx **pipe_ctx, * as well. */ for (i = 0; i < num_pipes; i++) { - pipe_ctx[i]->stream_res.tg->funcs->set_drr( - pipe_ctx[i]->stream_res.tg, ¶ms); - if (adjust.v_total_max != 0 && adjust.v_total_min != 0) - pipe_ctx[i]->stream_res.tg->funcs->set_static_screen_control( - pipe_ctx[i]->stream_res.tg, - event_triggers, num_frames); + if ((pipe_ctx[i]->stream_res.tg != NULL) && pipe_ctx[i]->stream_res.tg->funcs) { + if (pipe_ctx[i]->stream_res.tg->funcs->set_drr) + pipe_ctx[i]->stream_res.tg->funcs->set_drr( + pipe_ctx[i]->stream_res.tg, ¶ms); + if (adjust.v_total_max != 0 && adjust.v_total_min != 0) + if (pipe_ctx[i]->stream_res.tg->funcs->set_static_screen_control) + pipe_ctx[i]->stream_res.tg->funcs->set_static_screen_control( + pipe_ctx[i]->stream_res.tg, + event_triggers, num_frames); + } } } -- cgit v1.2.3 From aa4be41617a8034f73aa7fdaf7d8a0573a6a2b9d Mon Sep 17 00:00:00 2001 From: Evgenii Krasnikov Date: Wed, 26 Jan 2022 15:11:53 -0500 Subject: drm/amd/display: ensure PSR force_static flag can always be set [HOW&WHY] Make sure psr_force_static() can always be called regardless of psr_allow_active value. Reviewed-by: Harry Vanzylldejong Reviewed-by: Evgenii Krasnikov Reviewed-by: Nicholas Choi Acked-by: Pavle Kotarac Signed-off-by: Evgenii Krasnikov Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc_link.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index c1a727ff8b1d..b625b46ae96c 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -3092,6 +3092,10 @@ bool dc_link_set_psr_allow_active(struct dc_link *link, const bool *allow_active psr->funcs->psr_set_power_opt(psr, link->psr_settings.psr_power_opt, panel_inst); } + if (psr != NULL && link->psr_settings.psr_feature_enabled && + force_static && psr->funcs->psr_force_static) + psr->funcs->psr_force_static(psr, panel_inst); + /* Enable or Disable PSR */ if (allow_active && link->psr_settings.psr_allow_active != *allow_active) { link->psr_settings.psr_allow_active = *allow_active; @@ -3102,8 +3106,6 @@ bool dc_link_set_psr_allow_active(struct dc_link *link, const bool *allow_active #endif if (psr != NULL && link->psr_settings.psr_feature_enabled) { - if (force_static && psr->funcs->psr_force_static) - psr->funcs->psr_force_static(psr, panel_inst); psr->funcs->psr_enable(psr, link->psr_settings.psr_allow_active, wait, panel_inst); } else if ((dmcu != NULL && dmcu->funcs->is_dmcu_initialized(dmcu)) && link->psr_settings.psr_feature_enabled) -- cgit v1.2.3 From 382aceb305789d2e038baf2565c39e4117fa58d5 Mon Sep 17 00:00:00 2001 From: Charlene Liu Date: Wed, 30 Mar 2022 15:25:00 -0400 Subject: drm/amd/display: remove dtbclk_ss compensation for dcn316 [why] dcn316's dtbclk is from non_ss clock source. no compensation required here. Reviewed-by: Chris Park Acked-by: Pavle Kotarac Signed-off-by: Charlene Liu Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c | 2 +- drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c | 4 ++-- drivers/gpu/drm/amd/display/dc/dc.h | 2 +- 3 files changed, 4 insertions(+), 4 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c index dfba6138f538..26feefbb8990 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c @@ -374,7 +374,7 @@ void dce_clock_read_ss_info(struct clk_mgr_internal *clk_mgr_dce) clk_mgr_dce->dprefclk_ss_percentage = info.spread_spectrum_percentage; } - if (clk_mgr_dce->base.ctx->dc->debug.ignore_dpref_ss) + if (clk_mgr_dce->base.ctx->dc->config.ignore_dpref_ss) clk_mgr_dce->dprefclk_ss_percentage = 0; } } diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c index 702d00ce7da4..3121dd2d2a91 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c @@ -686,8 +686,8 @@ void dcn316_clk_mgr_construct( clk_mgr->base.base.dprefclk_khz = dcn316_smu_get_dpref_clk(&clk_mgr->base); clk_mgr->base.dccg->ref_dtbclk_khz = clk_mgr->base.base.dprefclk_khz; dce_clock_read_ss_info(&clk_mgr->base); - clk_mgr->base.dccg->ref_dtbclk_khz = - dce_adjust_dp_ref_freq_for_ss(&clk_mgr->base, clk_mgr->base.base.dprefclk_khz); + /*clk_mgr->base.dccg->ref_dtbclk_khz = + dce_adjust_dp_ref_freq_for_ss(&clk_mgr->base, clk_mgr->base.base.dprefclk_khz);*/ clk_mgr->base.base.bw_params = &dcn316_bw_params; diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 70cc21ec643a..e723553f9c5a 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -340,6 +340,7 @@ struct dc_config { bool is_asymmetric_memory; bool is_single_rank_dimm; bool use_pipe_ctx_sync_logic; + bool ignore_dpref_ss; }; enum visual_confirm { @@ -729,7 +730,6 @@ struct dc_debug_options { bool apply_vendor_specific_lttpr_wa; bool extended_blank_optimization; union aux_wake_wa_options aux_wake_wa; - bool ignore_dpref_ss; uint8_t psr_power_use_phy_fsm; }; -- cgit v1.2.3 From 593692d2de458a68c8e2326d48d1325d87947d90 Mon Sep 17 00:00:00 2001 From: Paul Hsieh Date: Wed, 16 Mar 2022 06:59:46 +0800 Subject: drm/amd/display: Power down hardware if timer not trigger [WHY] In headless systems, if SetMode/Power down timer is not called, hardware will not be powered down causing HW/SW discrepancies. Powering down hardware on SetPowerState to D3 will ensure SW/HW state is accurate. [HOW] 1. If PowerDownThread timer is not trigger but OS call SetPowerState to D3, power down hardware. 2. Update HDMI hang w/a to apply to all TMDS signals on headless system Reviewed-by: Martin Leung Acked-by: Pavle Kotarac Signed-off-by: Paul Hsieh Signed-off-by: Alex Deucher --- .../drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c | 17 +++++++++++--- .../amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c | 26 +++++++++++++++++++++- 2 files changed, 39 insertions(+), 4 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c index f4dee0e48a67..02943ca65807 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c @@ -88,11 +88,22 @@ static int rn_get_active_display_cnt_wa(struct dc *dc, struct dc_state *context) static void rn_set_low_power_state(struct clk_mgr *clk_mgr_base) { + int display_count; struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); + struct dc *dc = clk_mgr_base->ctx->dc; + struct dc_state *context = dc->current_state; + + if (clk_mgr_base->clks.pwr_state != DCN_PWR_STATE_LOW_POWER) { + + display_count = rn_get_active_display_cnt_wa(dc, context); - rn_vbios_smu_set_dcn_low_power_state(clk_mgr, DCN_PWR_STATE_LOW_POWER); - /* update power state */ - clk_mgr_base->clks.pwr_state = DCN_PWR_STATE_LOW_POWER; + /* if we can go lower, go lower */ + if (display_count == 0) { + rn_vbios_smu_set_dcn_low_power_state(clk_mgr, DCN_PWR_STATE_LOW_POWER); + /* update power state */ + clk_mgr_base->clks.pwr_state = DCN_PWR_STATE_LOW_POWER; + } + } } static void rn_update_clocks_update_dpp_dto(struct clk_mgr_internal *clk_mgr, diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c index 59fdd7f0d609..969b40250434 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_clk_mgr.c @@ -615,13 +615,37 @@ static void dcn31_clk_mgr_helper_populate_bw_params(struct clk_mgr_internal *clk } } +void dcn31_set_low_power_state(struct clk_mgr *clk_mgr_base) +{ + int display_count; + struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base); + struct dc *dc = clk_mgr_base->ctx->dc; + struct dc_state *context = dc->current_state; + + if (clk_mgr_base->clks.pwr_state != DCN_PWR_STATE_LOW_POWER) { + display_count = dcn31_get_active_display_cnt_wa(dc, context); + /* if we can go lower, go lower */ + if (display_count == 0) { + union display_idle_optimization_u idle_info = { 0 }; + + idle_info.idle_info.df_request_disabled = 1; + idle_info.idle_info.phy_ref_clk_off = 1; + idle_info.idle_info.s0i2_rdy = 1; + dcn31_smu_set_display_idle_optimization(clk_mgr, idle_info.data); + /* update power state */ + clk_mgr_base->clks.pwr_state = DCN_PWR_STATE_LOW_POWER; + } + } +} + static struct clk_mgr_funcs dcn31_funcs = { .get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz, .update_clocks = dcn31_update_clocks, .init_clocks = dcn31_init_clocks, .enable_pme_wa = dcn31_enable_pme_wa, .are_clock_states_equal = dcn31_are_clock_states_equal, - .notify_wm_ranges = dcn31_notify_wm_ranges + .notify_wm_ranges = dcn31_notify_wm_ranges, + .set_low_power_state = dcn31_set_low_power_state }; extern struct clk_mgr_funcs dcn3_fpga_funcs; -- cgit v1.2.3 From c1b972a18d05d007f0ddff31db2ff50790576e92 Mon Sep 17 00:00:00 2001 From: Oliver Logush Date: Tue, 22 Mar 2022 10:26:19 -0400 Subject: drm/amd/display: Insert pulling smu busy status before sending another request [why] Make sure smu is not busy before sending another request, this is to prevent stress failures from MS. [how] Check to make sure the SMU fw busy signal is cleared before sending another request Reviewed-by: Charlene Liu Reviewed-by: Nicholas Kazlauskas Acked-by: Pavle Kotarac Signed-off-by: Oliver Logush Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c index 8161a6ae410d..30c6f9cd717f 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c @@ -94,6 +94,9 @@ static int rn_vbios_smu_send_msg_with_param(struct clk_mgr_internal *clk_mgr, { uint32_t result; + result = rn_smu_wait_for_response(clk_mgr, 10, 200000); + ASSERT(result == VBIOSSMC_Result_OK); + /* First clear response register */ REG_WRITE(MP1_SMN_C2PMSG_91, VBIOSSMC_Status_BUSY); -- cgit v1.2.3 From 56b5b1c7b3d7319d4bf980cf0dd05d7c808c230d Mon Sep 17 00:00:00 2001 From: Oliver Logush Date: Tue, 29 Mar 2022 17:39:04 -0400 Subject: drm/amd/display: Always update DPPCLK_DTO when clock increase [why] Need to update the update_clock sequence to a fully tested sequence for dcn30 [how] Removed the check to see if clock is lowered Reviewed-by: Charlene Liu Acked-by: Pavle Kotarac Signed-off-by: Oliver Logush Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c | 3 +-- drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c | 3 +-- 2 files changed, 2 insertions(+), 4 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c index fbdd0a92d146..451e8d6cd8bd 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c @@ -157,8 +157,7 @@ static void dcn201_update_clocks(struct clk_mgr *clk_mgr_base, } else { if (update_dppclk || update_dispclk) dcn20_update_clocks_update_dentist(clk_mgr, context); - if (new_clocks->dppclk_khz >= dc->current_state->bw_ctx.bw.dcn.clk.dppclk_khz) - dcn20_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower); + dcn20_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower); } } } diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c index bc4ddc36fe58..f310b0d25a07 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c @@ -176,8 +176,7 @@ static void vg_update_clocks(struct clk_mgr *clk_mgr_base, if (update_dppclk || update_dispclk) dcn301_smu_set_dppclk(clk_mgr, clk_mgr_base->clks.dppclk_khz); // always update dtos unless clock is lowered and not safe to lower - if (new_clocks->dppclk_khz >= dc->current_state->bw_ctx.bw.dcn.clk.dppclk_khz) - dcn20_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower); + dcn20_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower); } } -- cgit v1.2.3 From 663d2daeaee6fa55530d8493979e9349070cd47c Mon Sep 17 00:00:00 2001 From: Duncan Ma Date: Thu, 31 Mar 2022 15:13:22 -0400 Subject: drm/amd/display: Add odm seamless boot support [WHY] Implement changes to transition from Pre-OS odm to Post-OS odm support. Seamless boot case is also considered. [HOW] Revised validation logic when marking for seamless boot. Init resources accordingly when Pre-OS has odm enabled. Reset odm and det size when transitioning Pre-OS odm to Post-OS non-odm to avoid corruption. Apply logic to set odm accordingly upon commit. Reviewed-by: Nicholas Kazlauskas Acked-by: Pavle Kotarac Signed-off-by: Duncan Ma Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc.c | 13 ++++ drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 82 ++++++++++++++-------- drivers/gpu/drm/amd/display/dc/dc.h | 1 + drivers/gpu/drm/amd/display/dc/dc_stream.h | 1 + .../drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 21 ++++++ drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c | 21 ++++++ drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.h | 2 + .../drm/amd/display/dc/inc/hw/timing_generator.h | 2 + 8 files changed, 115 insertions(+), 28 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index c436db416708..c2fcd67bcc4d 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -1569,11 +1569,24 @@ bool dc_validate_boot_timing(const struct dc *dc, if (dc_is_dp_signal(link->connector_signal)) { unsigned int pix_clk_100hz; + uint32_t numOdmPipes = 1; + uint32_t id_src[4] = {0}; dc->res_pool->dp_clock_source->funcs->get_pixel_clk_frequency_100hz( dc->res_pool->dp_clock_source, tg_inst, &pix_clk_100hz); + if (tg->funcs->get_optc_source) + tg->funcs->get_optc_source(tg, + &numOdmPipes, &id_src[0], &id_src[1]); + + if (numOdmPipes == 2) + pix_clk_100hz *= 2; + if (numOdmPipes == 4) + pix_clk_100hz *= 4; + + // Note: In rare cases, HW pixclk may differ from crtc's pixclk + // slightly due to rounding issues in 10 kHz units. if (crtc_timing->pix_clk_100hz != pix_clk_100hz) return false; diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c index f5777a71f2f1..f292303b75a5 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c @@ -2120,6 +2120,8 @@ static int acquire_resource_from_hw_enabled_state( { struct dc_link *link = stream->link; unsigned int i, inst, tg_inst = 0; + uint32_t numPipes = 1; + uint32_t id_src[4] = {0}; /* Check for enabled DIG to identify enabled display */ if (!link->link_enc->funcs->is_dig_enabled(link->link_enc)) @@ -2148,38 +2150,62 @@ static int acquire_resource_from_hw_enabled_state( if (!res_ctx->pipe_ctx[tg_inst].stream) { struct pipe_ctx *pipe_ctx = &res_ctx->pipe_ctx[tg_inst]; - pipe_ctx->stream_res.tg = pool->timing_generators[tg_inst]; - pipe_ctx->plane_res.mi = pool->mis[tg_inst]; - pipe_ctx->plane_res.hubp = pool->hubps[tg_inst]; - pipe_ctx->plane_res.ipp = pool->ipps[tg_inst]; - pipe_ctx->plane_res.xfm = pool->transforms[tg_inst]; - pipe_ctx->plane_res.dpp = pool->dpps[tg_inst]; - pipe_ctx->stream_res.opp = pool->opps[tg_inst]; - - if (pool->dpps[tg_inst]) { - pipe_ctx->plane_res.mpcc_inst = pool->dpps[tg_inst]->inst; - - // Read DPP->MPCC->OPP Pipe from HW State - if (pool->mpc->funcs->read_mpcc_state) { - struct mpcc_state s = {0}; - - pool->mpc->funcs->read_mpcc_state(pool->mpc, pipe_ctx->plane_res.mpcc_inst, &s); - - if (s.dpp_id < MAX_MPCC) - pool->mpc->mpcc_array[pipe_ctx->plane_res.mpcc_inst].dpp_id = s.dpp_id; - - if (s.bot_mpcc_id < MAX_MPCC) - pool->mpc->mpcc_array[pipe_ctx->plane_res.mpcc_inst].mpcc_bot = - &pool->mpc->mpcc_array[s.bot_mpcc_id]; + id_src[0] = tg_inst; + + if (pipe_ctx->stream_res.tg->funcs->get_optc_source) + pipe_ctx->stream_res.tg->funcs->get_optc_source(pipe_ctx->stream_res.tg, + &numPipes, &id_src[0], &id_src[1]); + + for (i = 0; i < numPipes; i++) { + //Check if src id invalid + if (id_src[i] == 0xf) + return -1; + + pipe_ctx->stream_res.tg = pool->timing_generators[tg_inst]; + pipe_ctx->plane_res.mi = pool->mis[id_src[i]]; + pipe_ctx->plane_res.hubp = pool->hubps[id_src[i]]; + pipe_ctx->plane_res.ipp = pool->ipps[id_src[i]]; + pipe_ctx->plane_res.xfm = pool->transforms[id_src[i]]; + pipe_ctx->plane_res.dpp = pool->dpps[id_src[i]]; + pipe_ctx->stream_res.opp = pool->opps[id_src[i]]; + + if (pool->dpps[id_src[i]]) { + pipe_ctx->plane_res.mpcc_inst = pool->dpps[id_src[i]]->inst; + + if (pool->mpc->funcs->read_mpcc_state) { + struct mpcc_state s = {0}; + pool->mpc->funcs->read_mpcc_state(pool->mpc, pipe_ctx->plane_res.mpcc_inst, &s); + if (s.dpp_id < MAX_MPCC) + pool->mpc->mpcc_array[pipe_ctx->plane_res.mpcc_inst].dpp_id = + s.dpp_id; + if (s.bot_mpcc_id < MAX_MPCC) + pool->mpc->mpcc_array[pipe_ctx->plane_res.mpcc_inst].mpcc_bot = + &pool->mpc->mpcc_array[s.bot_mpcc_id]; + if (s.opp_id < MAX_OPP) + pipe_ctx->stream_res.opp->mpc_tree_params.opp_id = s.opp_id; + } + } + pipe_ctx->pipe_idx = id_src[i]; - if (s.opp_id < MAX_OPP) - pipe_ctx->stream_res.opp->mpc_tree_params.opp_id = s.opp_id; + if (id_src[i] >= pool->timing_generator_count) { + id_src[i] = pool->timing_generator_count - 1; + pipe_ctx->stream_res.tg = pool->timing_generators[id_src[i]]; + pipe_ctx->stream_res.opp = pool->opps[id_src[i]]; } + + pipe_ctx->stream = stream; } - pipe_ctx->pipe_idx = tg_inst; - pipe_ctx->stream = stream; - return tg_inst; + if (numPipes == 2) { + stream->apply_boot_odm_mode = dm_odm_combine_policy_2to1; + res_ctx->pipe_ctx[id_src[0]].next_odm_pipe = &res_ctx->pipe_ctx[id_src[1]]; + res_ctx->pipe_ctx[id_src[0]].prev_odm_pipe = NULL; + res_ctx->pipe_ctx[id_src[1]].next_odm_pipe = NULL; + res_ctx->pipe_ctx[id_src[1]].prev_odm_pipe = &res_ctx->pipe_ctx[id_src[0]]; + } else + stream->apply_boot_odm_mode = dm_odm_combine_mode_disabled; + + return id_src[0]; } return -1; diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index e723553f9c5a..863d90bec61b 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -666,6 +666,7 @@ struct dc_debug_options { uint32_t edid_read_retry_times; bool remove_disconnect_edp; unsigned int force_odm_combine; //bit vector based on otg inst + unsigned int seamless_boot_odm_combine; #if defined(CONFIG_DRM_AMD_DC_DCN) unsigned int force_odm_combine_4to1; //bit vector based on otg inst bool disable_z9_mpc; diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h index c4168c11257c..580420c3eedc 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_stream.h +++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h @@ -246,6 +246,7 @@ struct dc_stream_state { bool apply_edp_fast_boot_optimization; bool apply_seamless_boot_optimization; + uint32_t apply_boot_odm_mode; uint32_t stream_id; diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index 6c69fd3bc77a..e02ac75afbf7 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c @@ -1259,6 +1259,7 @@ void dcn10_init_pipes(struct dc *dc, struct dc_state *context) { int i; struct dce_hwseq *hws = dc->hwseq; + struct hubbub *hubbub = dc->res_pool->hubbub; bool can_apply_seamless_boot = false; for (i = 0; i < context->stream_count; i++) { @@ -1294,6 +1295,21 @@ void dcn10_init_pipes(struct dc *dc, struct dc_state *context) } } + /* Reset det size */ + for (i = 0; i < dc->res_pool->pipe_count; i++) { + struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; + struct hubp *hubp = dc->res_pool->hubps[i]; + + /* Do not need to reset for seamless boot */ + if (pipe_ctx->stream != NULL && can_apply_seamless_boot) + continue; + + if (hubbub && hubp) { + if (hubbub->funcs->program_det_size) + hubbub->funcs->program_det_size(hubbub, hubp->inst, 0); + } + } + /* num_opp will be equal to number of mpcc */ for (i = 0; i < dc->res_pool->res_cap->num_opp; i++) { struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i]; @@ -1359,6 +1375,11 @@ void dcn10_init_pipes(struct dc *dc, struct dc_state *context) pipe_ctx->stream_res.tg = NULL; pipe_ctx->plane_res.hubp = NULL; + if (tg->funcs->is_tg_enabled(tg)) { + if (tg->funcs->init_odm) + tg->funcs->init_odm(tg); + } + tg->funcs->tg_init(tg); } diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c index 10f897b1cb63..c51f7dca94f8 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.c @@ -213,6 +213,26 @@ void optc31_set_drr( } } +void optc3_init_odm(struct timing_generator *optc) +{ + struct optc *optc1 = DCN10TG_FROM_TG(optc); + + REG_SET_5(OPTC_DATA_SOURCE_SELECT, 0, + OPTC_NUM_OF_INPUT_SEGMENT, 0, + OPTC_SEG0_SRC_SEL, optc->inst, + OPTC_SEG1_SRC_SEL, 0xf, + OPTC_SEG2_SRC_SEL, 0xf, + OPTC_SEG3_SRC_SEL, 0xf + ); + + REG_SET(OTG_H_TIMING_CNTL, 0, + OTG_H_TIMING_DIV_MODE, 0); + + REG_SET(OPTC_MEMORY_CONFIG, 0, + OPTC_MEM_SEL, 0); + optc1->opp_count = 1; +} + static struct timing_generator_funcs dcn31_tg_funcs = { .validate_timing = optc1_validate_timing, .program_timing = optc1_program_timing, @@ -272,6 +292,7 @@ static struct timing_generator_funcs dcn31_tg_funcs = { .program_manual_trigger = optc2_program_manual_trigger, .setup_manual_trigger = optc2_setup_manual_trigger, .get_hw_timing = optc1_get_hw_timing, + .init_odm = optc3_init_odm, }; void dcn31_timing_generator_init(struct optc *optc1) diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.h b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.h index a37b16040c1d..9e881f2ce74b 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.h +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_optc.h @@ -258,4 +258,6 @@ void dcn31_timing_generator_init(struct optc *optc1); void optc31_set_drr(struct timing_generator *optc, const struct drr_params *params); +void optc3_init_odm(struct timing_generator *optc); + #endif /* __DC_OPTC_DCN31_H__ */ diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h index 59a704781e34..554d2e33bd7f 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h @@ -310,6 +310,8 @@ struct timing_generator_funcs { uint32_t slave_pixel_clock_100Hz, uint8_t master_clock_divider, uint8_t slave_clock_divider); + + void (*init_odm)(struct timing_generator *tg); }; #endif -- cgit v1.2.3 From 30ffa74a04b6e48080f3181b29d2693b8486ca1e Mon Sep 17 00:00:00 2001 From: Jingwen Zhu Date: Fri, 1 Apr 2022 06:30:56 +0800 Subject: drm/amd/display: add DCN30 support FEC init [Why] FEC init used on DCN30. [How] Check fec active when HW init. Co-authored-by: Jingwen Zhu Reviewed-by: Wenjing Liu Acked-by: Pavle Kotarac Signed-off-by: Jingwen Zhu Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c index f61ec8763844..782b8db451b4 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c @@ -535,8 +535,12 @@ void dcn30_init_hw(struct dc *dc) /* Check for enabled DIG to identify enabled display */ if (link->link_enc->funcs->is_dig_enabled && - link->link_enc->funcs->is_dig_enabled(link->link_enc)) + link->link_enc->funcs->is_dig_enabled(link->link_enc)) { link->link_status.link_active = true; + if (link->link_enc->funcs->fec_is_active && + link->link_enc->funcs->fec_is_active(link->link_enc)) + link->fec_state = dc_link_fec_enabled; + } } /* Power gate DSCs */ -- cgit v1.2.3 From 19a2e1e36a2b14d7549a6d9584be131f4286f757 Mon Sep 17 00:00:00 2001 From: Angus Wang Date: Thu, 31 Mar 2022 09:33:10 -0400 Subject: drm/amd/display: Fix inconsistent timestamp type [WHY] An unsigned int timestamp variable is assigned with an unsigned long long value. Also, the assignment directly converts the tick value to us without using built-in get elapsed time function. [HOW] Cast the assigned value correctly and also use built-in function to get the timestamp in the unit we want. v2: squash in 64 bit division fix Reviewed-by: Aric Cyr Acked-by: Pavle Kotarac Signed-off-by: Angus Wang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/modules/freesync/freesync.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c index d2d76ce56f89..03fa63d56fa6 100644 --- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c +++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c @@ -1230,6 +1230,7 @@ void mod_freesync_handle_v_update(struct mod_freesync *mod_freesync, { struct core_freesync *core_freesync = NULL; unsigned int cur_timestamp_in_us; + unsigned long long cur_tick; if ((mod_freesync == NULL) || (stream == NULL) || (in_out_vrr == NULL)) return; @@ -1239,7 +1240,9 @@ void mod_freesync_handle_v_update(struct mod_freesync *mod_freesync, if (in_out_vrr->supported == false) return; - cur_timestamp_in_us = div_u64(dm_get_timestamp(core_freesync->dc->ctx), 10); + cur_tick = dm_get_timestamp(core_freesync->dc->ctx); + cur_timestamp_in_us = (unsigned int) + div_u64(dm_get_elapse_time_in_ns(core_freesync->dc->ctx, cur_tick, 0), 1000); in_out_vrr->flip_interval.vsyncs_between_flip++; in_out_vrr->flip_interval.v_update_timestamp_in_us = cur_timestamp_in_us; -- cgit v1.2.3 From 3b69632c85ebb687c35e03be4d4ef70e384fd990 Mon Sep 17 00:00:00 2001 From: Dillon Varone Date: Wed, 30 Mar 2022 18:42:24 -0400 Subject: drm/amd/display: Add dtb clock to dc_clocks Reviewed-by: Ariel Bernstein Acked-by: Pavle Kotarac Signed-off-by: Dillon Varone Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dc.h | 1 + 1 file changed, 1 insertion(+) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 863d90bec61b..544d58260269 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -417,6 +417,7 @@ struct dc_clocks { #if defined(CONFIG_DRM_AMD_DC_DCN) enum dcn_zstate_support_state zstate_support; bool dtbclk_en; + int dtbclk_khz; #endif enum dcn_pwr_state pwr_state; /* -- cgit v1.2.3 From 634c67ee33a7bfb574aba1796d89493ff6614ab7 Mon Sep 17 00:00:00 2001 From: hersen wu Date: Fri, 1 Apr 2022 15:25:41 -0400 Subject: drm/amd/display: expose skip_detection_link_training to debugfs [Why] within dc link detecion, dp link training will be executed for external sst dp. for debug purpose, we may need skip dp link training. [How] expose dc debug option to skip_detection_link_training to debugfs Reviewed-by: Roman Li Acked-by: Pavle Kotarac Signed-off-by: hersen wu Signed-off-by: Alex Deucher --- .../drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c | 37 ++++++++++++++++++++++ 1 file changed, 37 insertions(+) (limited to 'drivers/gpu') 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 da17ece1a2c5..188039f14544 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 @@ -3491,6 +3491,40 @@ DEFINE_SHOW_ATTRIBUTE(mst_topo); DEFINE_DEBUGFS_ATTRIBUTE(visual_confirm_fops, visual_confirm_get, visual_confirm_set, "%llu\n"); + +/* + * Sets the DC skip_detection_link_training debug option from the given string. + * Example usage: echo 1 > /sys/kernel/debug/dri/0/amdgpu_skip_detection_link_training + */ +static int skip_detection_link_training_set(void *data, u64 val) +{ + struct amdgpu_device *adev = data; + + if (val == 0) + adev->dm.dc->debug.skip_detection_link_training = false; + else + adev->dm.dc->debug.skip_detection_link_training = true; + + return 0; +} + +/* + * Reads the DC skip_detection_link_training debug option value into the given buffer. + * Example usage: cat /sys/kernel/debug/dri/0/amdgpu_dm_skip_detection_link_training + */ +static int skip_detection_link_training_get(void *data, u64 *val) +{ + struct amdgpu_device *adev = data; + + *val = adev->dm.dc->debug.skip_detection_link_training; + + return 0; +} + +DEFINE_DEBUGFS_ATTRIBUTE(skip_detection_link_training_fops, + skip_detection_link_training_get, + skip_detection_link_training_set, "%llu\n"); + /* * Dumps the DCC_EN bit for each pipe. * Example usage: cat /sys/kernel/debug/dri/0/amdgpu_dm_dcc_en @@ -3584,6 +3618,9 @@ void dtn_debugfs_init(struct amdgpu_device *adev) debugfs_create_file_unsafe("amdgpu_dm_visual_confirm", 0644, root, adev, &visual_confirm_fops); + debugfs_create_file_unsafe("amdgpu_dm_skip_detection_link_training", 0644, root, adev, + &skip_detection_link_training_fops); + debugfs_create_file_unsafe("amdgpu_dm_dmub_tracebuffer", 0644, root, adev, &dmub_tracebuffer_fops); -- cgit v1.2.3 From 34d292d57973dd432e93425de33f8f120cfbdab3 Mon Sep 17 00:00:00 2001 From: Felix Kuehling Date: Thu, 7 Apr 2022 22:08:03 -0400 Subject: drm/amdkfd: Asynchronously free events The synchronize_rcu call in destroy_events can take several ms, which noticeably slows down applications destroying many events. Use kfree_rcu to free the event structure asynchronously and eliminate the synchronize_rcu call in the user thread. Signed-off-by: Felix Kuehling Reviewed-by: Philip Yang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_events.c | 3 +-- drivers/gpu/drm/amd/amdkfd/kfd_events.h | 2 ++ 2 files changed, 3 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c index 0fef24b0b915..75847c5d5957 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c @@ -262,8 +262,7 @@ static void destroy_event(struct kfd_process *p, struct kfd_event *ev) p->signal_event_count--; idr_remove(&p->event_idr, ev->event_id); - synchronize_rcu(); - kfree(ev); + kfree_rcu(ev, rcu); } static void destroy_events(struct kfd_process *p) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.h b/drivers/gpu/drm/amd/amdkfd/kfd_events.h index 55d376f56021..1c62c8dd6460 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.h @@ -70,6 +70,8 @@ struct kfd_event { struct kfd_hsa_memory_exception_data memory_exception_data; struct kfd_hsa_hw_exception_data hw_exception_data; }; + + struct rcu_head rcu; /* for asynchronous kfree_rcu */ }; #define KFD_EVENT_TIMEOUT_IMMEDIATE 0 -- cgit v1.2.3 From e9b1bfa548be678fcfd706291637b51f35ee92b8 Mon Sep 17 00:00:00 2001 From: Dillon Varone Date: Fri, 1 Apr 2022 16:45:40 -0400 Subject: drm/amd/display: Select correct DTO source [WHY&HOW] Change criteria for setting DTO source value, and always set it regardless of the signal type. Reviewed-by: Ariel Bernstein Acked-by: Pavle Kotarac Signed-off-by: Dillon Varone Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c index c36f8e829344..760653e2b607 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c @@ -985,7 +985,7 @@ static bool dcn31_program_pix_clk( struct bp_pixel_clock_parameters bp_pc_params = {0}; enum transmitter_color_depth bp_pc_colour_depth = TRANSMITTER_COLOR_DEPTH_24; // For these signal types Driver to program DP_DTO without calling VBIOS Command table - if (dc_is_dp_signal(pix_clk_params->signal_type)) { + if (dc_is_dp_signal(pix_clk_params->signal_type) || dc_is_virtual_signal(pix_clk_params->signal_type)) { if (e) { /* Set DTO values: phase = target clock, modulo = reference clock*/ REG_WRITE(PHASE[inst], e->target_pixel_rate_khz * e->mult_factor); -- cgit v1.2.3 From c2e4b0a97cc2ed0590fade2aebc4d912dd868ff3 Mon Sep 17 00:00:00 2001 From: Aric Cyr Date: Mon, 4 Apr 2022 00:22:58 -0400 Subject: drm/amd/display: 3.2.181 Title: DC Patches Apri 6, 2022 This DC patchset brings improvements in multiple areas. In summary, we highlight: *Disabling Z10 on DCN31 *Fix issue breaking 32bit Linux build *Fix inconsistent timestamp type *Add DCN30 support FEC init *Fix crash on setting VRR with no display connected *Disable FEC if DSC not supported for EDP *Add odm seamless boot support *Select correct DTO source *Power down hardware if timer not trigger Acked-by: Pavle Kotarac Signed-off-by: Aric Cyr Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dc.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 544d58260269..2f0c436dae4c 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -47,7 +47,7 @@ struct aux_payload; struct set_config_cmd_payload; struct dmub_notification; -#define DC_VER "3.2.180" +#define DC_VER "3.2.181" #define MAX_SURFACES 3 #define MAX_PLANES 6 -- cgit v1.2.3 From 9dd941015718c1387d6f7f61bd4394d38d273c50 Mon Sep 17 00:00:00 2001 From: David Zhang Date: Wed, 23 Mar 2022 23:11:48 -0400 Subject: drm/amd/display: implement shared PSR-SU sink validation helper [why] creating a generic helper for AMD specific PSR-SU sink validation. Moving the function to the power module to reference it across all OS. [how] - drop PSRSU specific sink validation helper and move to power module by reading PSR version and other PSR caps - call the new helper from linux DM (amdgpu_dm_psr) Acked-by: Pavle Kotarac Acked-by: Tom Chung Signed-off-by: David Zhang Signed-off-by: Alex Deucher --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c | 9 ++++- .../drm/amd/display/modules/power/power_helpers.c | 38 ++++++++++++++++++++++ .../drm/amd/display/modules/power/power_helpers.h | 1 + 3 files changed, 47 insertions(+), 1 deletion(-) (limited to 'drivers/gpu') 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 0c923a90615c..13b1751e69bf 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 @@ -27,6 +27,7 @@ #include "dc.h" #include "dm_helpers.h" #include "amdgpu_dm.h" +#include "modules/power/power_helpers.h" #ifdef CONFIG_DRM_AMD_DC_DCN static bool link_supports_psrsu(struct dc_link *link) @@ -39,6 +40,9 @@ static bool link_supports_psrsu(struct dc_link *link) if (dc->ctx->dce_version < DCN_VERSION_3_1) return false; + if (!is_psr_su_specific_panel(link)) + return false; + if (!link->dpcd_caps.alpm_caps.bits.AUX_WAKE_ALPM_CAP || !link->dpcd_caps.psr_info.psr_dpcd_caps.bits.Y_COORDINATE_REQUIRED) return false; @@ -79,7 +83,10 @@ void amdgpu_dm_set_psr_caps(struct dc_link *link) link->psr_settings.psr_feature_enabled = true; } - DRM_INFO("PSR support:%d\n", link->psr_settings.psr_feature_enabled); + DRM_INFO("PSR support %d, DC PSR ver %d, sink PSR ver %d\n", + link->psr_settings.psr_feature_enabled, + link->psr_settings.psr_version, + link->dpcd_caps.psr_info.psr_version); } diff --git a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c index 2b00f334e93d..97928d4c3b9a 100644 --- a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c +++ b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c @@ -784,3 +784,41 @@ bool dmcu_load_iram(struct dmcu *dmcu, return result; } +/* + * is_psr_su_specific_panel() - check if sink is AMD vendor-specific PSR-SU + * supported eDP device. + * + * @link: dc link pointer + * + * Return: true if AMDGPU vendor specific PSR-SU eDP panel + */ +bool is_psr_su_specific_panel(struct dc_link *link) +{ + if (link->dpcd_caps.edp_rev >= DP_EDP_14) { + if (link->dpcd_caps.psr_info.psr_version >= DP_PSR2_WITH_Y_COORD_ET_SUPPORTED) + return true; + /* + * Some panels will report PSR capabilities over additional DPCD bits. + * Such panels are approved despite reporting only PSR v3, as long as + * the additional bits are reported. + */ + if (link->dpcd_caps.psr_info.psr_version < DP_PSR2_WITH_Y_COORD_IS_SUPPORTED) + return false; + + if (link->dpcd_caps.sink_dev_id == DP_BRANCH_DEVICE_ID_001CF8) { + /* + * FIXME: + * This is the temporary workaround to disable PSRSU when system turned on + * DSC function on the sepcific sink. Once the PSRSU + DSC is fixed, this + * condition should be removed. + */ + if (link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_SUPPORT) + return false; + + if (link->dpcd_caps.psr_info.force_psrsu_cap == 0x1) + return true; + } + } + + return false; +} diff --git a/drivers/gpu/drm/amd/display/modules/power/power_helpers.h b/drivers/gpu/drm/amd/display/modules/power/power_helpers.h index 2a9f8e2d8080..1a634d8c78c5 100644 --- a/drivers/gpu/drm/amd/display/modules/power/power_helpers.h +++ b/drivers/gpu/drm/amd/display/modules/power/power_helpers.h @@ -52,4 +52,5 @@ bool dmub_init_abm_config(struct resource_pool *res_pool, struct dmcu_iram_parameters params, unsigned int inst); +bool is_psr_su_specific_panel(struct dc_link *link); #endif /* MODULES_POWER_POWER_HELPERS_H_ */ -- cgit v1.2.3 From a5b7994306f66a62acb48734b5577809a87a1cf2 Mon Sep 17 00:00:00 2001 From: Wenjing Liu Date: Fri, 25 Mar 2022 16:12:14 -0400 Subject: drm/amd/display: extract update stream allocation to link_hwss [why] Extract update stream allocation table into link hwss as part of the link hwss refactor work. Reviewed-by: George Shen Reviewed-by: Fangzhi Zuo Acked-by: Pavle Kotarac Signed-off-by: Wenjing Liu Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/core/dc_link.c | 95 ++++++---------------- drivers/gpu/drm/amd/display/dc/inc/link_hwss.h | 8 +- .../gpu/drm/amd/display/dc/link/link_hwss_dio.c | 11 +++ .../gpu/drm/amd/display/dc/link/link_hwss_dpia.c | 27 ++++++ .../gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c | 10 +++ 5 files changed, 78 insertions(+), 73 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index b625b46ae96c..faab1460d0b5 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -3487,8 +3487,6 @@ static enum dc_status dc_link_update_sst_payload(struct pipe_ctx *pipe_ctx, { struct dc_stream_state *stream = pipe_ctx->stream; struct dc_link *link = stream->link; - struct hpo_dp_link_encoder *hpo_dp_link_encoder = pipe_ctx->link_res.hpo_dp_link_enc; - struct hpo_dp_stream_encoder *hpo_dp_stream_encoder = pipe_ctx->stream_res.hpo_dp_stream_enc; struct link_mst_stream_allocation_table proposed_table = {0}; struct fixed31_32 avg_time_slots_per_mtp; const struct dc_link_settings empty_link_settings = {0}; @@ -3522,7 +3520,7 @@ static enum dc_status dc_link_update_sst_payload(struct pipe_ctx *pipe_ctx, pipe_ctx->pipe_idx); } - proposed_table.stream_allocations[0].hpo_dp_stream_enc = hpo_dp_stream_encoder; + proposed_table.stream_allocations[0].hpo_dp_stream_enc = pipe_ctx->stream_res.hpo_dp_stream_enc; ASSERT(proposed_table.stream_count == 1); @@ -3535,8 +3533,7 @@ static enum dc_status dc_link_update_sst_payload(struct pipe_ctx *pipe_ctx, proposed_table.stream_allocations[0].slot_count); /* program DP source TX for payload */ - hpo_dp_link_encoder->funcs->update_stream_allocation_table( - hpo_dp_link_encoder, + link_hwss->ext.update_stream_allocation_table(link, &pipe_ctx->link_res, &proposed_table); /* poll for ACT handled */ @@ -3573,8 +3570,6 @@ enum dc_status dc_link_allocate_mst_payload(struct pipe_ctx *pipe_ctx) { struct dc_stream_state *stream = pipe_ctx->stream; struct dc_link *link = stream->link; - struct link_encoder *link_encoder = NULL; - struct hpo_dp_link_encoder *hpo_dp_link_encoder = pipe_ctx->link_res.hpo_dp_link_enc; struct dp_mst_stream_allocation_table proposed_table = {0}; struct fixed31_32 avg_time_slots_per_mtp; struct fixed31_32 pbn; @@ -3584,9 +3579,6 @@ enum dc_status dc_link_allocate_mst_payload(struct pipe_ctx *pipe_ctx) const struct link_hwss *link_hwss = get_link_hwss(link, &pipe_ctx->link_res); DC_LOGGER_INIT(link->ctx->logger); - link_encoder = link_enc_cfg_get_link_enc(link); - ASSERT(link_encoder); - /* enable_link_dp_mst already check link->enabled_stream_count * and stream is in link->stream[]. This is called during set mode, * stream_enc is available. @@ -3631,37 +3623,17 @@ enum dc_status dc_link_allocate_mst_payload(struct pipe_ctx *pipe_ctx) ASSERT(proposed_table.stream_count > 0); - if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA) { - static enum dc_status status; - uint8_t mst_alloc_slots = 0, prev_mst_slots_in_use = 0xFF; - - for (i = 0; i < link->mst_stream_alloc_table.stream_count; i++) - mst_alloc_slots += link->mst_stream_alloc_table.stream_allocations[i].slot_count; - - status = dc_process_dmub_set_mst_slots(link->dc, link->link_index, - mst_alloc_slots, &prev_mst_slots_in_use); - ASSERT(status == DC_OK); - DC_LOG_MST("dpia : status[%d]: alloc_slots[%d]: used_slots[%d]\n", - status, mst_alloc_slots, prev_mst_slots_in_use); - } - /* program DP source TX for payload */ - switch (dp_get_link_encoding_format(&link->cur_link_settings)) { - case DP_8b_10b_ENCODING: - link_encoder->funcs->update_mst_stream_allocation_table( - link_encoder, - &link->mst_stream_alloc_table); - break; - case DP_128b_132b_ENCODING: - hpo_dp_link_encoder->funcs->update_stream_allocation_table( - hpo_dp_link_encoder, - &link->mst_stream_alloc_table); - break; - case DP_UNKNOWN_ENCODING: + if (link_hwss->ext.update_stream_allocation_table == NULL || + dp_get_link_encoding_format(&link->cur_link_settings) == DP_UNKNOWN_ENCODING) { DC_LOG_ERROR("Failure: unknown encoding format\n"); return DC_ERROR_UNEXPECTED; } + link_hwss->ext.update_stream_allocation_table(link, + &pipe_ctx->link_res, + &link->mst_stream_alloc_table); + /* send down message */ ret = dm_helpers_dp_mst_poll_for_allocation_change_trigger( stream->ctx, @@ -3703,7 +3675,6 @@ enum dc_status dc_link_reduce_mst_payload(struct pipe_ctx *pipe_ctx, uint32_t bw struct fixed31_32 avg_time_slots_per_mtp; struct fixed31_32 pbn; struct fixed31_32 pbn_per_slot; - struct link_encoder *link_encoder = link->link_enc; struct dp_mst_stream_allocation_table proposed_table = {0}; uint8_t i; enum act_return_status ret; @@ -3767,8 +3738,13 @@ enum dc_status dc_link_reduce_mst_payload(struct pipe_ctx *pipe_ctx, uint32_t bw ASSERT(proposed_table.stream_count > 0); /* update mst stream allocation table hardware state */ - link_encoder->funcs->update_mst_stream_allocation_table( - link_encoder, + if (link_hwss->ext.update_stream_allocation_table == NULL || + dp_get_link_encoding_format(&link->cur_link_settings) == DP_UNKNOWN_ENCODING) { + DC_LOG_ERROR("Failure: unknown encoding format\n"); + return DC_ERROR_UNEXPECTED; + } + + link_hwss->ext.update_stream_allocation_table(link, &pipe_ctx->link_res, &link->mst_stream_alloc_table); /* poll for immediate branch device ACT handled */ @@ -3863,8 +3839,6 @@ static enum dc_status deallocate_mst_payload(struct pipe_ctx *pipe_ctx) { struct dc_stream_state *stream = pipe_ctx->stream; struct dc_link *link = stream->link; - struct link_encoder *link_encoder = NULL; - struct hpo_dp_link_encoder *hpo_dp_link_encoder = pipe_ctx->link_res.hpo_dp_link_enc; struct dp_mst_stream_allocation_table proposed_table = {0}; struct fixed31_32 avg_time_slots_per_mtp = dc_fixpt_from_int(0); int i; @@ -3873,9 +3847,6 @@ static enum dc_status deallocate_mst_payload(struct pipe_ctx *pipe_ctx) const struct dc_link_settings empty_link_settings = {0}; DC_LOGGER_INIT(link->ctx->logger); - link_encoder = link_enc_cfg_get_link_enc(link); - ASSERT(link_encoder); - /* deallocate_mst_payload is called before disable link. When mode or * disable/enable monitor, new stream is created which is not in link * stream[] yet. For this, payload is not allocated yet, so de-alloc @@ -3933,36 +3904,16 @@ static enum dc_status deallocate_mst_payload(struct pipe_ctx *pipe_ctx) link->mst_stream_alloc_table.stream_allocations[i].slot_count); } - if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA) { - enum dc_status status; - uint8_t mst_alloc_slots = 0, prev_mst_slots_in_use = 0xFF; - - for (i = 0; i < link->mst_stream_alloc_table.stream_count; i++) - mst_alloc_slots += link->mst_stream_alloc_table.stream_allocations[i].slot_count; - - status = dc_process_dmub_set_mst_slots(link->dc, link->link_index, - mst_alloc_slots, &prev_mst_slots_in_use); - ASSERT(status != DC_NOT_SUPPORTED); - DC_LOG_MST("dpia : status[%d]: alloc_slots[%d]: used_slots[%d]\n", - status, mst_alloc_slots, prev_mst_slots_in_use); - } - - switch (dp_get_link_encoding_format(&link->cur_link_settings)) { - case DP_8b_10b_ENCODING: - link_encoder->funcs->update_mst_stream_allocation_table( - link_encoder, - &link->mst_stream_alloc_table); - break; - case DP_128b_132b_ENCODING: - hpo_dp_link_encoder->funcs->update_stream_allocation_table( - hpo_dp_link_encoder, - &link->mst_stream_alloc_table); - break; - case DP_UNKNOWN_ENCODING: + /* update mst stream allocation table hardware state */ + if (link_hwss->ext.update_stream_allocation_table == NULL || + dp_get_link_encoding_format(&link->cur_link_settings) == DP_UNKNOWN_ENCODING) { DC_LOG_DEBUG("Unknown encoding format\n"); return DC_ERROR_UNEXPECTED; } + link_hwss->ext.update_stream_allocation_table(link, &pipe_ctx->link_res, + &link->mst_stream_alloc_table); + if (mst_mode) { dm_helpers_dp_mst_poll_for_allocation_change_trigger( stream->ctx, @@ -4109,8 +4060,8 @@ static void fpga_dp_hpo_enable_link_and_stream(struct dc_state *state, struct pi proposed_table.stream_allocations[0].hpo_dp_stream_enc = pipe_ctx->stream_res.hpo_dp_stream_enc; } - pipe_ctx->link_res.hpo_dp_link_enc->funcs->update_stream_allocation_table( - pipe_ctx->link_res.hpo_dp_link_enc, + link_hwss->ext.update_stream_allocation_table(stream->link, + &pipe_ctx->link_res, &proposed_table); if (link_hwss->ext.set_throttled_vcp_size) diff --git a/drivers/gpu/drm/amd/display/dc/inc/link_hwss.h b/drivers/gpu/drm/amd/display/dc/inc/link_hwss.h index e2f854464c05..e6c49ef8b584 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/link_hwss.h +++ b/drivers/gpu/drm/amd/display/dc/inc/link_hwss.h @@ -37,9 +37,12 @@ struct dc_link; struct link_resource; struct pipe_ctx; struct encoder_set_dp_phy_pattern_param; +struct link_mst_stream_allocation_table; struct link_hwss_ext { - /* function pointers below require check for NULL at all time + /* function pointers below may require to check for NULL if caller + * considers missing implementation as expected in some cases or none + * critical to be investigated immediately * ********************************************************************* */ void (*set_hblank_min_symbol_width)(struct pipe_ctx *pipe_ctx, @@ -62,6 +65,9 @@ struct link_hwss_ext { const struct link_resource *link_res, const struct dc_link_settings *link_settings, const struct dc_lane_settings lane_settings[LANE_COUNT_DP_MAX]); + void (*update_stream_allocation_table)(struct dc_link *link, + const struct link_resource *link_res, + const struct link_mst_stream_allocation_table *table); }; struct link_hwss { diff --git a/drivers/gpu/drm/amd/display/dc/link/link_hwss_dio.c b/drivers/gpu/drm/amd/display/dc/link/link_hwss_dio.c index e557c92acddc..776e822abcbb 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_hwss_dio.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_hwss_dio.c @@ -153,6 +153,16 @@ void set_dio_dp_lane_settings(struct dc_link *link, link_enc->funcs->dp_set_lane_settings(link_enc, link_settings, lane_settings); } +static void update_dio_stream_allocation_table(struct dc_link *link, + const struct link_resource *link_res, + const struct link_mst_stream_allocation_table *table) +{ + struct link_encoder *link_enc = link_enc_cfg_get_link_enc(link); + + ASSERT(link_enc); + link_enc->funcs->update_mst_stream_allocation_table(link_enc, table); +} + static const struct link_hwss dio_link_hwss = { .setup_stream_encoder = setup_dio_stream_encoder, .reset_stream_encoder = reset_dio_stream_encoder, @@ -163,6 +173,7 @@ static const struct link_hwss dio_link_hwss = { .disable_dp_link_output = disable_dio_dp_link_output, .set_dp_link_test_pattern = set_dio_dp_link_test_pattern, .set_dp_lane_settings = set_dio_dp_lane_settings, + .update_stream_allocation_table = update_dio_stream_allocation_table, }, }; diff --git a/drivers/gpu/drm/amd/display/dc/link/link_hwss_dpia.c b/drivers/gpu/drm/amd/display/dc/link/link_hwss_dpia.c index f90596f37cd4..89d4e8159138 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_hwss_dpia.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_hwss_dpia.c @@ -25,6 +25,32 @@ #include "link_hwss_dpia.h" #include "core_types.h" #include "link_hwss_dio.h" +#include "link_enc_cfg.h" + +#define DC_LOGGER_INIT(logger) + +static void update_dpia_stream_allocation_table(struct dc_link *link, + const struct link_resource *link_res, + const struct link_mst_stream_allocation_table *table) +{ + struct link_encoder *link_enc = link_enc_cfg_get_link_enc(link); + static enum dc_status status; + uint8_t mst_alloc_slots = 0, prev_mst_slots_in_use = 0xFF; + int i; + DC_LOGGER_INIT(link->ctx->logger); + + for (i = 0; i < table->stream_count; i++) + mst_alloc_slots += table->stream_allocations[i].slot_count; + + status = dc_process_dmub_set_mst_slots(link->dc, link->link_index, + mst_alloc_slots, &prev_mst_slots_in_use); + ASSERT(status == DC_OK); + DC_LOG_MST("dpia : status[%d]: alloc_slots[%d]: used_slots[%d]\n", + status, mst_alloc_slots, prev_mst_slots_in_use); + + ASSERT(link_enc); + link_enc->funcs->update_mst_stream_allocation_table(link_enc, table); +} static const struct link_hwss dpia_link_hwss = { .setup_stream_encoder = setup_dio_stream_encoder, @@ -36,6 +62,7 @@ static const struct link_hwss dpia_link_hwss = { .disable_dp_link_output = disable_dio_dp_link_output, .set_dp_link_test_pattern = set_dio_dp_link_test_pattern, .set_dp_lane_settings = set_dio_dp_lane_settings, + .update_stream_allocation_table = update_dpia_stream_allocation_table, }, }; diff --git a/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c b/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c index 311e2e2687f0..87972dc8443d 100644 --- a/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c +++ b/drivers/gpu/drm/amd/display/dc/link/link_hwss_hpo_dp.c @@ -244,6 +244,15 @@ static void set_hpo_dp_lane_settings(struct dc_link *link, lane_settings[0].FFE_PRESET.raw); } +static void update_hpo_dp_stream_allocation_table(struct dc_link *link, + const struct link_resource *link_res, + const struct link_mst_stream_allocation_table *table) +{ + link_res->hpo_dp_link_enc->funcs->update_stream_allocation_table( + link_res->hpo_dp_link_enc, + table); +} + static const struct link_hwss hpo_dp_link_hwss = { .setup_stream_encoder = setup_hpo_dp_stream_encoder, .reset_stream_encoder = reset_hpo_dp_stream_encoder, @@ -255,6 +264,7 @@ static const struct link_hwss hpo_dp_link_hwss = { .disable_dp_link_output = disable_hpo_dp_link_output, .set_dp_link_test_pattern = set_hpo_dp_link_test_pattern, .set_dp_lane_settings = set_hpo_dp_lane_settings, + .update_stream_allocation_table = update_hpo_dp_stream_allocation_table, }, }; -- cgit v1.2.3 From 3925f9b4fed1b51bfe17770a63609a7a0f6e4808 Mon Sep 17 00:00:00 2001 From: Lang Yu Date: Tue, 12 Apr 2022 10:41:53 +0800 Subject: drm/amdkfd: shrink bitmap size in struct svm_validate_context A MAX_GPU_INSTANCE bits bitmap will suffice. Signed-off-by: Lang Yu Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c index 8b6adc142e25..459fa07a3bcc 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c @@ -1370,7 +1370,7 @@ struct svm_validate_context { struct kfd_process *process; struct svm_range *prange; bool intr; - unsigned long bitmap[MAX_GPU_INSTANCE]; + DECLARE_BITMAP(bitmap, MAX_GPU_INSTANCE); struct ttm_validate_buffer tv[MAX_GPU_INSTANCE]; struct list_head validate_list; struct ww_acquire_ctx ticket; -- cgit v1.2.3 From eb85fc238970afe4a309bb307bdeecc2938ba7b2 Mon Sep 17 00:00:00 2001 From: Yongqiang Sun Date: Fri, 8 Apr 2022 10:53:14 -0400 Subject: drm/amd/amdgpu: Not request init data for MS_HYPERV with vega10 MS_HYPERV with vega10 doesn't have the interface to process request init data msg. Check hypervisor type to not send the request for MS_HYPERV. Signed-off-by: Yongqiang Sun Reviewed-by: Alice Wong Acked-by: Alex Deucher Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index 9aa355a5ac3c..ab95beac6307 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -23,6 +23,10 @@ #include +#ifdef CONFIG_X86 +#include +#endif + #include #include "amdgpu.h" @@ -723,8 +727,12 @@ void amdgpu_detect_virtualization(struct amdgpu_device *adev) break; case CHIP_VEGA10: soc15_set_virt_ops(adev); - /* send a dummy GPU_INIT_DATA request to host on vega10 */ - amdgpu_virt_request_init_data(adev); +#ifdef CONFIG_X86 + /* not send GPU_INIT_DATA with MS_HYPERV*/ + if (!hypervisor_is_type(X86_HYPER_MS_HYPERV)) +#endif + /* send a dummy GPU_INIT_DATA request to host on vega10 */ + amdgpu_virt_request_init_data(adev); break; case CHIP_VEGA20: case CHIP_ARCTURUS: -- cgit v1.2.3 From 46d18d510d78318c4aa5aaeff66782f1ec42c2ec Mon Sep 17 00:00:00 2001 From: Mukul Joshi Date: Wed, 6 Apr 2022 20:07:37 -0400 Subject: drm/amdkfd: Cleanup IO links during KFD device removal Currently, the IO-links to the device being removed from topology, are not cleared. As a result, there would be dangling links left in the KFD topology. This patch aims to fix the following: 1. Cleanup all IO links to the device being removed. 2. Ensure that node numbering in sysfs and nodes proximity domain values are consistent after the device is removed: a. Adding a device and removing a GPU device are made mutually exclusive. b. The global proximity domain counter is no longer required to be an atomic counter. A normal 32-bit counter can be used instead. 3. Update generation_count to let user-mode know that topology has changed due to device removal. CC: Shuotao Xu Reviewed-by: Shuotao Xu Reviewed-by: Felix Kuehling Signed-off-by: Mukul Joshi Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_crat.c | 4 +- drivers/gpu/drm/amd/amdkfd/kfd_priv.h | 2 + drivers/gpu/drm/amd/amdkfd/kfd_topology.c | 83 +++++++++++++++++++++++++++---- 3 files changed, 78 insertions(+), 11 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c index 1eaabd2cb41b..afc8a7fcdad8 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c @@ -1056,7 +1056,7 @@ static int kfd_parse_subtype_iolink(struct crat_subtype_iolink *iolink, * table, add corresponded reversed direction link now. */ if (props && (iolink->flags & CRAT_IOLINK_FLAGS_BI_DIRECTIONAL)) { - to_dev = kfd_topology_device_by_proximity_domain(id_to); + to_dev = kfd_topology_device_by_proximity_domain_no_lock(id_to); if (!to_dev) return -ENODEV; /* same everything but the other direction */ @@ -2225,7 +2225,7 @@ static int kfd_create_vcrat_image_gpu(void *pcrat_image, */ if (kdev->hive_id) { for (nid = 0; nid < proximity_domain; ++nid) { - peer_dev = kfd_topology_device_by_proximity_domain(nid); + peer_dev = kfd_topology_device_by_proximity_domain_no_lock(nid); if (!peer_dev->gpu) continue; if (peer_dev->gpu->hive_id != kdev->hive_id) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index e1b7e6afa920..8a43def1f638 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -1016,6 +1016,8 @@ int kfd_topology_add_device(struct kfd_dev *gpu); int kfd_topology_remove_device(struct kfd_dev *gpu); struct kfd_topology_device *kfd_topology_device_by_proximity_domain( uint32_t proximity_domain); +struct kfd_topology_device *kfd_topology_device_by_proximity_domain_no_lock( + uint32_t proximity_domain); struct kfd_topology_device *kfd_topology_device_by_id(uint32_t gpu_id); struct kfd_dev *kfd_device_by_id(uint32_t gpu_id); struct kfd_dev *kfd_device_by_pci_dev(const struct pci_dev *pdev); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c index 3bdcae239bc0..8b7710b4d3ed 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c @@ -46,22 +46,32 @@ static struct list_head topology_device_list; static struct kfd_system_properties sys_props; static DECLARE_RWSEM(topology_lock); -static atomic_t topology_crat_proximity_domain; +static uint32_t topology_crat_proximity_domain; -struct kfd_topology_device *kfd_topology_device_by_proximity_domain( +struct kfd_topology_device *kfd_topology_device_by_proximity_domain_no_lock( uint32_t proximity_domain) { struct kfd_topology_device *top_dev; struct kfd_topology_device *device = NULL; - down_read(&topology_lock); - list_for_each_entry(top_dev, &topology_device_list, list) if (top_dev->proximity_domain == proximity_domain) { device = top_dev; break; } + return device; +} + +struct kfd_topology_device *kfd_topology_device_by_proximity_domain( + uint32_t proximity_domain) +{ + struct kfd_topology_device *device = NULL; + + down_read(&topology_lock); + + device = kfd_topology_device_by_proximity_domain_no_lock( + proximity_domain); up_read(&topology_lock); return device; @@ -1060,7 +1070,7 @@ int kfd_topology_init(void) down_write(&topology_lock); kfd_topology_update_device_list(&temp_topology_device_list, &topology_device_list); - atomic_set(&topology_crat_proximity_domain, sys_props.num_devices-1); + topology_crat_proximity_domain = sys_props.num_devices-1; ret = kfd_topology_update_sysfs(); up_write(&topology_lock); @@ -1295,8 +1305,6 @@ int kfd_topology_add_device(struct kfd_dev *gpu) pr_debug("Adding new GPU (ID: 0x%x) to topology\n", gpu_id); - proximity_domain = atomic_inc_return(&topology_crat_proximity_domain); - /* Include the CPU in xGMI hive if xGMI connected by assigning it the hive ID. */ if (gpu->hive_id && gpu->adev->gmc.xgmi.connected_to_cpu) { struct kfd_topology_device *top_dev; @@ -1321,12 +1329,16 @@ int kfd_topology_add_device(struct kfd_dev *gpu) */ dev = kfd_assign_gpu(gpu); if (!dev) { + down_write(&topology_lock); + proximity_domain = ++topology_crat_proximity_domain; + res = kfd_create_crat_image_virtual(&crat_image, &image_size, COMPUTE_UNIT_GPU, gpu, proximity_domain); if (res) { pr_err("Error creating VCRAT for GPU (ID: 0x%x)\n", gpu_id); + topology_crat_proximity_domain--; return res; } res = kfd_parse_crat_table(crat_image, @@ -1335,10 +1347,10 @@ int kfd_topology_add_device(struct kfd_dev *gpu) if (res) { pr_err("Error parsing VCRAT for GPU (ID: 0x%x)\n", gpu_id); + topology_crat_proximity_domain--; goto err; } - down_write(&topology_lock); kfd_topology_update_device_list(&temp_topology_device_list, &topology_device_list); @@ -1485,25 +1497,78 @@ err: return res; } +/** + * kfd_topology_update_io_links() - Update IO links after device removal. + * @proximity_domain: Proximity domain value of the dev being removed. + * + * The topology list currently is arranged in increasing order of + * proximity domain. + * + * Two things need to be done when a device is removed: + * 1. All the IO links to this device need to be removed. + * 2. All nodes after the current device node need to move + * up once this device node is removed from the topology + * list. As a result, the proximity domain values for + * all nodes after the node being deleted reduce by 1. + * This would also cause the proximity domain values for + * io links to be updated based on new proximity domain + * values. + * + * Context: The caller must hold write topology_lock. + */ +static void kfd_topology_update_io_links(int proximity_domain) +{ + struct kfd_topology_device *dev; + struct kfd_iolink_properties *iolink, *tmp; + + list_for_each_entry(dev, &topology_device_list, list) { + if (dev->proximity_domain > proximity_domain) + dev->proximity_domain--; + + list_for_each_entry_safe(iolink, tmp, &dev->io_link_props, list) { + /* + * If there is an io link to the dev being deleted + * then remove that IO link also. + */ + if (iolink->node_to == proximity_domain) { + list_del(&iolink->list); + dev->io_link_count--; + dev->node_props.io_links_count--; + } else if (iolink->node_from > proximity_domain) { + iolink->node_from--; + } else if (iolink->node_to > proximity_domain) { + iolink->node_to--; + } + } + + } +} + int kfd_topology_remove_device(struct kfd_dev *gpu) { struct kfd_topology_device *dev, *tmp; uint32_t gpu_id; int res = -ENODEV; + int i = 0; down_write(&topology_lock); - list_for_each_entry_safe(dev, tmp, &topology_device_list, list) + list_for_each_entry_safe(dev, tmp, &topology_device_list, list) { if (dev->gpu == gpu) { gpu_id = dev->gpu_id; kfd_remove_sysfs_node_entry(dev); kfd_release_topology_device(dev); sys_props.num_devices--; + kfd_topology_update_io_links(i); + topology_crat_proximity_domain = sys_props.num_devices-1; + sys_props.generation_count++; res = 0; if (kfd_topology_update_sysfs() < 0) kfd_topology_release_sysfs(); break; } + i++; + } up_write(&topology_lock); -- cgit v1.2.3 From abb5bc59490067b88ed006cdb58f3ff1d12057cc Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Wed, 13 Apr 2022 10:36:40 +0300 Subject: drm/amdkfd: potential NULL dereference in kfd_set/reset_event() If lookup_event_by_id() returns a NULL "ev" pointer then the spin_lock(&ev->lock) will crash. This was detected by Smatch: drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_events.c:644 kfd_set_event() error: we previously assumed 'ev' could be null (see line 639) Fixes: 5273e82c5f47 ("drm/amdkfd: Improve concurrency of event handling") Signed-off-by: Dan Carpenter Reviewed-by: Felix Kuehling Signed-off-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_events.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c index 75847c5d5957..e449bc70bf08 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c @@ -633,14 +633,19 @@ int kfd_set_event(struct kfd_process *p, uint32_t event_id) rcu_read_lock(); ev = lookup_event_by_id(p, event_id); + if (!ev) { + ret = -EINVAL; + goto unlock_rcu; + } spin_lock(&ev->lock); - if (ev && event_can_be_cpu_signaled(ev)) + if (event_can_be_cpu_signaled(ev)) set_event(ev); else ret = -EINVAL; spin_unlock(&ev->lock); +unlock_rcu: rcu_read_unlock(); return ret; } @@ -659,14 +664,19 @@ int kfd_reset_event(struct kfd_process *p, uint32_t event_id) rcu_read_lock(); ev = lookup_event_by_id(p, event_id); + if (!ev) { + ret = -EINVAL; + goto unlock_rcu; + } spin_lock(&ev->lock); - if (ev && event_can_be_cpu_signaled(ev)) + if (event_can_be_cpu_signaled(ev)) reset_event(ev); else ret = -EINVAL; spin_unlock(&ev->lock); +unlock_rcu: rcu_read_unlock(); return ret; -- cgit v1.2.3 From 7c703a7d3f2b50a6187267420a4d3d7e62fa3206 Mon Sep 17 00:00:00 2001 From: xinhui pan Date: Tue, 12 Apr 2022 19:52:16 +0800 Subject: drm/amdgpu: Fix one use-after-free of VM MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit VM might already be freed when amdgpu_vm_tlb_seq_cb() is called. We see the calltrace below. Fix it by keeping the last flush fence around and wait for it to signal BUG kmalloc-4k (Not tainted): Poison overwritten 0xffff9c88630414e8-0xffff9c88630414e8 @offset=5352. First byte 0x6c instead of 0x6b Allocated in amdgpu_driver_open_kms+0x9d/0x360 [amdgpu] age=44 cpu=0 pid=2343 __slab_alloc.isra.0+0x4f/0x90 kmem_cache_alloc_trace+0x6b8/0x7a0 amdgpu_driver_open_kms+0x9d/0x360 [amdgpu] drm_file_alloc+0x222/0x3e0 [drm] drm_open+0x11d/0x410 [drm] Freed in amdgpu_driver_postclose_kms+0x3e9/0x550 [amdgpu] age=22 cpu=1 pid=2485 kfree+0x4a2/0x580 amdgpu_driver_postclose_kms+0x3e9/0x550 [amdgpu] drm_file_free+0x24e/0x3c0 [drm] drm_close_helper.isra.0+0x90/0xb0 [drm] drm_release+0x97/0x1a0 [drm] __fput+0xb6/0x280 ____fput+0xe/0x10 task_work_run+0x64/0xb0 Suggested-by: Christian König Signed-off-by: xinhui pan Reviewed-by: Christian König Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 18 +++++++++++++++--- drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h | 1 + 2 files changed, 16 insertions(+), 3 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 383242952a20..4736174f5e4d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -932,10 +932,14 @@ int amdgpu_vm_update_range(struct amdgpu_device *adev, struct amdgpu_vm *vm, if (flush_tlb || params.table_freed) { tlb_cb->vm = vm; - if (!fence || !*fence || - dma_fence_add_callback(*fence, &tlb_cb->cb, - amdgpu_vm_tlb_seq_cb)) + if (fence && *fence && + !dma_fence_add_callback(*fence, &tlb_cb->cb, + amdgpu_vm_tlb_seq_cb)) { + dma_fence_put(vm->last_tlb_flush); + vm->last_tlb_flush = dma_fence_get(*fence); + } else { amdgpu_vm_tlb_seq_cb(NULL, &tlb_cb->cb); + } tlb_cb = NULL; } @@ -2094,6 +2098,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm) vm->update_funcs = &amdgpu_vm_sdma_funcs; vm->last_update = NULL; vm->last_unlocked = dma_fence_get_stub(); + vm->last_tlb_flush = dma_fence_get_stub(); mutex_init(&vm->eviction_lock); vm->evicting = false; @@ -2132,6 +2137,7 @@ error_free_root: vm->root.bo = NULL; error_free_delayed: + dma_fence_put(vm->last_tlb_flush); dma_fence_put(vm->last_unlocked); drm_sched_entity_destroy(&vm->delayed); @@ -2248,6 +2254,7 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) struct amdgpu_bo_va_mapping *mapping, *tmp; bool prt_fini_needed = !!adev->gmc.gmc_funcs->set_prt; struct amdgpu_bo *root; + unsigned long flags; int i; amdgpu_amdkfd_gpuvm_destroy_cb(adev, vm); @@ -2257,6 +2264,11 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) amdgpu_vm_set_pasid(adev, vm, 0); dma_fence_wait(vm->last_unlocked, false); dma_fence_put(vm->last_unlocked); + dma_fence_wait(vm->last_tlb_flush, false); + /* Make sure that all fence callbacks have completed */ + spin_lock_irqsave(vm->last_tlb_flush->lock, flags); + spin_unlock_irqrestore(vm->last_tlb_flush->lock, flags); + dma_fence_put(vm->last_tlb_flush); list_for_each_entry_safe(mapping, tmp, &vm->freed, list) { if (mapping->flags & AMDGPU_PTE_PRT && prt_fini_needed) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index 1a814fbffff8..6b06a214f05f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -286,6 +286,7 @@ struct amdgpu_vm { /* Last finished delayed update */ atomic64_t tlb_seq; + struct dma_fence *last_tlb_flush; /* Last unlocked submission to the scheduler entities */ struct dma_fence *last_unlocked; -- cgit v1.2.3 From 9e12b5947bbaa74db32e24d77ebf3bd71015970c Mon Sep 17 00:00:00 2001 From: Rodrigo Siqueira Date: Tue, 12 Apr 2022 14:32:51 -0400 Subject: Revert "drm/amd/display: Pass HostVM enable flag into DCN3.1 DML" This reverts commit 863fa85e6a01c8b239009825dd9de1f64d7d020a. While we were testing DCN3.1 with a hub, we noticed that only one of 2 connected displays lights up when using some specific display resolution. In summary, this was the setup: 1. Displays: * Sharp LQ156M1JW26 (eDP): 1080@240 * BENQ SW320 (DP): 4k@60 * BENQ EX3203R (DP): 4k@60 2. Hub: Club3D CSV-7300 3. ASIC: DCN3.1 After bisecting this issue, we figured out the commit mentioned above introduced this issue. We are investigating why this patch introduced this regression, but we need to revert it for now. Cc: Harry Wentland Cc: Mark Broadworth Cc: Michael Strauss Acked-by: Alex Deucher Signed-off-by: Rodrigo Siqueira Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c | 1 - 1 file changed, 1 deletion(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c index 6cc580be7c79..5b3f0c2dfb55 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_resource.c @@ -1668,7 +1668,6 @@ int dcn31_populate_dml_pipes_from_context( pipes[pipe_cnt].pipe.src.immediate_flip = true; pipes[pipe_cnt].pipe.src.unbounded_req_mode = false; - pipes[pipe_cnt].pipe.src.hostvm = dc->res_pool->hubbub->riommu_active; pipes[pipe_cnt].pipe.src.gpuvm = true; pipes[pipe_cnt].pipe.src.dcc_fraction_of_zs_req_luma = 0; pipes[pipe_cnt].pipe.src.dcc_fraction_of_zs_req_chroma = 0; -- cgit v1.2.3 From 250e64a3f06c42e993e7b8c6f727d43b5b9f8c96 Mon Sep 17 00:00:00 2001 From: Felix Kuehling Date: Tue, 12 Apr 2022 16:24:49 -0400 Subject: drm/amdkfd: fix race condition in kfd_wait_on_events Add the waiters to the wait queue during initialization, while holding the event spinlock. Otherwise the waiter will not get activated if the event signals before being added to the wait queue. Signed-off-by: Felix Kuehling Reviewed-by: Philip Yang Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_events.c | 26 +++++--------------------- 1 file changed, 5 insertions(+), 21 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c index e449bc70bf08..6e5e8d637f48 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c @@ -780,7 +780,7 @@ static struct kfd_event_waiter *alloc_event_waiters(uint32_t num_events) return event_waiters; } -static int init_event_waiter_get_status(struct kfd_process *p, +static int init_event_waiter(struct kfd_process *p, struct kfd_event_waiter *waiter, uint32_t event_id) { @@ -793,25 +793,13 @@ static int init_event_waiter_get_status(struct kfd_process *p, waiter->event = ev; waiter->activated = ev->signaled; ev->signaled = ev->signaled && !ev->auto_reset; + if (!waiter->activated) + add_wait_queue(&ev->wq, &waiter->wait); spin_unlock(&ev->lock); return 0; } -static void init_event_waiter_add_to_waitlist(struct kfd_event_waiter *waiter) -{ - struct kfd_event *ev = waiter->event; - - /* Only add to the wait list if we actually need to - * wait on this event. - */ - if (!waiter->activated) { - spin_lock(&ev->lock); - add_wait_queue(&ev->wq, &waiter->wait); - spin_unlock(&ev->lock); - } -} - /* test_event_condition - Test condition of events being waited for * @all: Return completion only if all events have signaled * @num_events: Number of events to wait for @@ -941,8 +929,8 @@ int kfd_wait_on_events(struct kfd_process *p, goto out_unlock; } - ret = init_event_waiter_get_status(p, &event_waiters[i], - event_data.event_id); + ret = init_event_waiter(p, &event_waiters[i], + event_data.event_id); if (ret) goto out_unlock; } @@ -960,10 +948,6 @@ int kfd_wait_on_events(struct kfd_process *p, goto out_unlock; } - /* Add to wait lists if we need to wait. */ - for (i = 0; i < num_events; i++) - init_event_waiter_add_to_waitlist(&event_waiters[i]); - mutex_unlock(&p->event_mutex); while (true) { -- cgit v1.2.3 From d68cf992ded575928cf4ddf7c64faff0d8dcce14 Mon Sep 17 00:00:00 2001 From: Gavin Wan Date: Wed, 13 Apr 2022 11:09:22 -0400 Subject: drm/amd/amdgpu: Remove static from variable in RLCG Reg RW [why] These static variables save the RLC Scratch registers address. When we install multiple GPUs (for example: XGMI setting) and multiple GPUs call the function at same time. The RLC Scratch registers address are changed each other. Then it caused reading/writing from/to wrong GPU. [how] Removed the static from the variables. The variables are on the stack. Fixes: 5d447e29670148 ("drm/amdgpu: add helper for rlcg indirect reg access") Reviewed-by: Alex Deucher Signed-off-by: Gavin Wan Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'drivers/gpu') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index ab95beac6307..ea92edcc0432 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -870,11 +870,11 @@ static u32 amdgpu_virt_rlcg_reg_rw(struct amdgpu_device *adev, u32 offset, u32 v uint32_t timeout = 50000; uint32_t i, tmp; uint32_t ret = 0; - static void *scratch_reg0; - static void *scratch_reg1; - static void *scratch_reg2; - static void *scratch_reg3; - static void *spare_int; + void *scratch_reg0; + void *scratch_reg1; + void *scratch_reg2; + void *scratch_reg3; + void *spare_int; if (!adev->gfx.rlc.rlcg_reg_access_supported) { dev_err(adev->dev, -- cgit v1.2.3