drm/amdgpu: add shadow bo support V2

[karo-tx-linux.git] / drivers / gpu / drm / amd / amdgpu / amdgpu_object.c

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 6f0873c75a25650feb9c65330f1a3b133ab64f50..278017d3dc3fe45e6fc0681211157f099e53bb14 100644 (file)
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -211,6 +211,69 @@ static void amdgpu_fill_placement_to_bo(struct amdgpu_bo *bo,
        bo->placement.busy_placement = bo->placements;
 }
 
+/**
+ * amdgpu_bo_create_kernel - create BO for kernel use
+ *
+ * @adev: amdgpu device object
+ * @size: size for the new BO
+ * @align: alignment for the new BO
+ * @domain: where to place it
+ * @bo_ptr: resulting BO
+ * @gpu_addr: GPU addr of the pinned BO
+ * @cpu_addr: optional CPU address mapping
+ *
+ * Allocates and pins a BO for kernel internal use.
+ *
+ * Returns 0 on success, negative error code otherwise.
+ */
+int amdgpu_bo_create_kernel(struct amdgpu_device *adev,
+                           unsigned long size, int align,
+                           u32 domain, struct amdgpu_bo **bo_ptr,
+                           u64 *gpu_addr, void **cpu_addr)
+{
+       int r;
+
+       r = amdgpu_bo_create(adev, size, align, true, domain,
+                            AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
+                            NULL, NULL, bo_ptr);
+       if (r) {
+               dev_err(adev->dev, "(%d) failed to allocate kernel bo\n", r);
+               return r;
+       }
+
+       r = amdgpu_bo_reserve(*bo_ptr, false);
+       if (r) {
+               dev_err(adev->dev, "(%d) failed to reserve kernel bo\n", r);
+               goto error_free;
+       }
+
+       r = amdgpu_bo_pin(*bo_ptr, domain, gpu_addr);
+       if (r) {
+               dev_err(adev->dev, "(%d) kernel bo pin failed\n", r);
+               goto error_unreserve;
+       }
+
+       if (cpu_addr) {
+               r = amdgpu_bo_kmap(*bo_ptr, cpu_addr);
+               if (r) {
+                       dev_err(adev->dev, "(%d) kernel bo map failed\n", r);
+                       goto error_unreserve;
+               }
+       }
+
+       amdgpu_bo_unreserve(*bo_ptr);
+
+       return 0;
+
+error_unreserve:
+       amdgpu_bo_unreserve(*bo_ptr);
+
+error_free:
+       amdgpu_bo_unref(bo_ptr);
+
+       return r;
+}
+
 int amdgpu_bo_create_restricted(struct amdgpu_device *adev,
                                unsigned long size, int byte_align,
                                bool kernel, u32 domain, u64 flags,
@@ -277,11 +340,75 @@ int amdgpu_bo_create_restricted(struct amdgpu_device *adev,
        if (unlikely(r != 0)) {
                return r;
        }
+
+       if (flags & AMDGPU_GEM_CREATE_VRAM_CLEARED &&
+           bo->tbo.mem.placement & TTM_PL_FLAG_VRAM) {
+               struct fence *fence;
+
+               if (adev->mman.buffer_funcs_ring == NULL ||
+                  !adev->mman.buffer_funcs_ring->ready) {
+                       r = -EBUSY;
+                       goto fail_free;
+               }
+
+               r = amdgpu_bo_reserve(bo, false);
+               if (unlikely(r != 0))
+                       goto fail_free;
+
+               amdgpu_ttm_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_VRAM);
+               r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
+               if (unlikely(r != 0))
+                       goto fail_unreserve;
+
+               amdgpu_fill_buffer(bo, 0, bo->tbo.resv, &fence);
+               amdgpu_bo_fence(bo, fence, false);
+               amdgpu_bo_unreserve(bo);
+               fence_put(bo->tbo.moving);
+               bo->tbo.moving = fence_get(fence);
+               fence_put(fence);
+       }
        *bo_ptr = bo;
 
        trace_amdgpu_bo_create(bo);
 
        return 0;
+
+fail_unreserve:
+       amdgpu_bo_unreserve(bo);
+fail_free:
+       amdgpu_bo_unref(&bo);
+       return r;
+}
+
+static int amdgpu_bo_create_shadow(struct amdgpu_device *adev,
+                                  unsigned long size, int byte_align,
+                                  struct amdgpu_bo *bo)
+{
+       struct ttm_placement placement = {0};
+       struct ttm_place placements[AMDGPU_GEM_DOMAIN_MAX + 1];
+       int r;
+
+       if (bo->shadow)
+               return 0;
+
+       bo->flags |= AMDGPU_GEM_CREATE_SHADOW;
+       memset(&placements, 0,
+              (AMDGPU_GEM_DOMAIN_MAX + 1) * sizeof(struct ttm_place));
+
+       amdgpu_ttm_placement_init(adev, &placement,
+                                 placements, AMDGPU_GEM_DOMAIN_GTT,
+                                 AMDGPU_GEM_CREATE_CPU_GTT_USWC);
+
+       r = amdgpu_bo_create_restricted(adev, size, byte_align, true,
+                                       AMDGPU_GEM_DOMAIN_GTT,
+                                       AMDGPU_GEM_CREATE_CPU_GTT_USWC,
+                                       NULL, &placement,
+                                       bo->tbo.resv,
+                                       &bo->shadow);
+       if (!r)
+               bo->shadow->parent = amdgpu_bo_ref(bo);
+
+       return r;
 }
 
 int amdgpu_bo_create(struct amdgpu_device *adev,
@@ -293,6 +420,7 @@ int amdgpu_bo_create(struct amdgpu_device *adev,
 {
        struct ttm_placement placement = {0};
        struct ttm_place placements[AMDGPU_GEM_DOMAIN_MAX + 1];
+       int r;
 
        memset(&placements, 0,
               (AMDGPU_GEM_DOMAIN_MAX + 1) * sizeof(struct ttm_place));
@@ -300,9 +428,19 @@ int amdgpu_bo_create(struct amdgpu_device *adev,
        amdgpu_ttm_placement_init(adev, &placement,
                                  placements, domain, flags);
 
-       return amdgpu_bo_create_restricted(adev, size, byte_align, kernel,
-                                          domain, flags, sg, &placement,
-                                          resv, bo_ptr);
+       r = amdgpu_bo_create_restricted(adev, size, byte_align, kernel,
+                                       domain, flags, sg, &placement,
+                                       resv, bo_ptr);
+       if (r)
+               return r;
+
+       if (flags & AMDGPU_GEM_CREATE_SHADOW) {
+               r = amdgpu_bo_create_shadow(adev, size, byte_align, (*bo_ptr));
+               if (r)
+                       amdgpu_bo_unref(bo_ptr);
+       }
+
+       return r;
 }
 
 int amdgpu_bo_kmap(struct amdgpu_bo *bo, void **ptr)
@@ -674,3 +812,21 @@ void amdgpu_bo_fence(struct amdgpu_bo *bo, struct fence *fence,
        else
                reservation_object_add_excl_fence(resv, fence);
 }
+
+/**
+ * amdgpu_bo_gpu_offset - return GPU offset of bo
+ * @bo:        amdgpu object for which we query the offset
+ *
+ * Returns current GPU offset of the object.
+ *
+ * Note: object should either be pinned or reserved when calling this
+ * function, it might be useful to add check for this for debugging.
+ */
+u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo)
+{
+       WARN_ON_ONCE(bo->tbo.mem.mem_type == TTM_PL_SYSTEM);
+       WARN_ON_ONCE(!ww_mutex_is_locked(&bo->tbo.resv->lock) &&
+                    !bo->pin_count);
+
+       return bo->tbo.offset;
+}