1 files changed, 130 insertions, 122 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
index 489041df1f45..a59cd47aa6c1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dma_buf.c
@@ -34,27 +34,12 @@
 #include "amdgpu.h"
 #include "amdgpu_display.h"
 #include "amdgpu_gem.h"
+#include "amdgpu_dma_buf.h"
 #include <drm/amdgpu_drm.h>
 #include <linux/dma-buf.h>
 #include <linux/dma-fence-array.h>
 
 /**
- * amdgpu_gem_prime_get_sg_table - &drm_driver.gem_prime_get_sg_table
- * implementation
- * @obj: GEM buffer object (BO)
- *
- * Returns:
- * A scatter/gather table for the pinned pages of the BO's memory.
- */
-struct sg_table *amdgpu_gem_prime_get_sg_table(struct drm_gem_object *obj)
-{
-	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
-	int npages = bo->tbo.num_pages;
-
-	return drm_prime_pages_to_sg(bo->tbo.ttm->pages, npages);
-}
-
-/**
  * amdgpu_gem_prime_vmap - &dma_buf_ops.vmap implementation
  * @obj: GEM BO
  *
@@ -137,23 +122,23 @@ int amdgpu_gem_prime_mmap(struct drm_gem_object *obj,
 }
 
 static int
-__reservation_object_make_exclusive(struct reservation_object *obj)
+__dma_resv_make_exclusive(struct dma_resv *obj)
 {
 	struct dma_fence **fences;
 	unsigned int count;
 	int r;
 
-	if (!reservation_object_get_list(obj)) /* no shared fences to convert */
+	if (!dma_resv_get_list(obj)) /* no shared fences to convert */
 		return 0;
 
-	r = reservation_object_get_fences_rcu(obj, NULL, &count, &fences);
+	r = dma_resv_get_fences_rcu(obj, NULL, &count, &fences);
 	if (r)
 		return r;
 
 	if (count == 0) {
 		/* Now that was unexpected. */
 	} else if (count == 1) {
-		reservation_object_add_excl_fence(obj, fences[0]);
+		dma_resv_add_excl_fence(obj, fences[0]);
 		dma_fence_put(fences[0]);
 		kfree(fences);
 	} else {
@@ -165,7 +150,7 @@ __reservation_object_make_exclusive(struct reservation_object *obj)
 		if (!array)
 			goto err_fences_put;
 
-		reservation_object_add_excl_fence(obj, &array->base);
+		dma_resv_add_excl_fence(obj, &array->base);
 		dma_fence_put(&array->base);
 	}
 
@@ -179,106 +164,126 @@ err_fences_put:
 }
 
 /**
- * amdgpu_dma_buf_map_attach - &dma_buf_ops.attach implementation
- * @dma_buf: Shared DMA buffer
- * @attach: DMA-buf attachment
+ * amdgpu_dma_buf_attach - &dma_buf_ops.attach implementation
  *
- * Makes sure that the shared DMA buffer can be accessed by the target device.
- * For now, simply pins it to the GTT domain, where it should be accessible by
- * all DMA devices.
+ * @dmabuf: DMA-buf where we attach to
+ * @attach: attachment to add
  *
- * Returns:
- * 0 on success or a negative error code on failure.
+ * Add the attachment as user to the exported DMA-buf.
  */
-static int amdgpu_dma_buf_map_attach(struct dma_buf *dma_buf,
-				     struct dma_buf_attachment *attach)
+static int amdgpu_dma_buf_attach(struct dma_buf *dmabuf,
+				 struct dma_buf_attachment *attach)
 {
-	struct drm_gem_object *obj = dma_buf->priv;
+	struct drm_gem_object *obj = dmabuf->priv;
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-	long r;
+	int r;
 
-	r = drm_gem_map_attach(dma_buf, attach);
-	if (r)
-		return r;
+	if (attach->dev->driver == adev->dev->driver)
+		return 0;
 
 	r = amdgpu_bo_reserve(bo, false);
 	if (unlikely(r != 0))
-		goto error_detach;
-
-
-	if (attach->dev->driver != adev->dev->driver) {
-		/*
-		 * We only create shared fences for internal use, but importers
-		 * of the dmabuf rely on exclusive fences for implicitly
-		 * tracking write hazards. As any of the current fences may
-		 * correspond to a write, we need to convert all existing
-		 * fences on the reservation object into a single exclusive
-		 * fence.
-		 */
-		r = __reservation_object_make_exclusive(bo->tbo.resv);
-		if (r)
-			goto error_unreserve;
-	}
+		return r;
 
-	/* pin buffer into GTT */
-	r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT);
+	/*
+	 * We only create shared fences for internal use, but importers
+	 * of the dmabuf rely on exclusive fences for implicitly
+	 * tracking write hazards. As any of the current fences may
+	 * correspond to a write, we need to convert all existing
+	 * fences on the reservation object into a single exclusive
+	 * fence.
+	 */
+	r = __dma_resv_make_exclusive(bo->tbo.base.resv);
 	if (r)
-		goto error_unreserve;
-
-	if (attach->dev->driver != adev->dev->driver)
-		bo->prime_shared_count++;
+		return r;
 
-error_unreserve:
+	bo->prime_shared_count++;
 	amdgpu_bo_unreserve(bo);
+	return 0;
+}
 
-error_detach:
-	if (r)
-		drm_gem_map_detach(dma_buf, attach);
-	return r;
+/**
+ * amdgpu_dma_buf_detach - &dma_buf_ops.detach implementation
+ *
+ * @dmabuf: DMA-buf where we remove the attachment from
+ * @attach: the attachment to remove
+ *
+ * Called when an attachment is removed from the DMA-buf.
+ */
+static void amdgpu_dma_buf_detach(struct dma_buf *dmabuf,
+				  struct dma_buf_attachment *attach)
+{
+	struct drm_gem_object *obj = dmabuf->priv;
+	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
+	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
+
+	if (attach->dev->driver != adev->dev->driver && bo->prime_shared_count)
+		bo->prime_shared_count--;
 }
 
 /**
- * amdgpu_dma_buf_map_detach - &dma_buf_ops.detach implementation
- * @dma_buf: Shared DMA buffer
+ * amdgpu_dma_buf_map - &dma_buf_ops.map_dma_buf implementation
  * @attach: DMA-buf attachment
+ * @dir: DMA direction
  *
- * This is called when a shared DMA buffer no longer needs to be accessible by
- * another device. For now, simply unpins the buffer from GTT.
+ * Makes sure that the shared DMA buffer can be accessed by the target device.
+ * For now, simply pins it to the GTT domain, where it should be accessible by
+ * all DMA devices.
+ *
+ * Returns:
+ * sg_table filled with the DMA addresses to use or ERR_PRT with negative error
+ * code.
  */
-static void amdgpu_dma_buf_map_detach(struct dma_buf *dma_buf,
-				      struct dma_buf_attachment *attach)
+static struct sg_table *amdgpu_dma_buf_map(struct dma_buf_attachment *attach,
+					   enum dma_data_direction dir)
 {
+	struct dma_buf *dma_buf = attach->dmabuf;
 	struct drm_gem_object *obj = dma_buf->priv;
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
-	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-	int ret = 0;
+	struct sg_table *sgt;
+	long r;
 
-	ret = amdgpu_bo_reserve(bo, true);
-	if (unlikely(ret != 0))
-		goto error;
+	r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT);
+	if (r)
+		return ERR_PTR(r);
 
-	amdgpu_bo_unpin(bo);
-	if (attach->dev->driver != adev->dev->driver && bo->prime_shared_count)
-		bo->prime_shared_count--;
-	amdgpu_bo_unreserve(bo);
+	sgt = drm_prime_pages_to_sg(bo->tbo.ttm->pages, bo->tbo.num_pages);
+	if (IS_ERR(sgt))
+		return sgt;
 
-error:
-	drm_gem_map_detach(dma_buf, attach);
+	if (!dma_map_sg_attrs(attach->dev, sgt->sgl, sgt->nents, dir,
+			      DMA_ATTR_SKIP_CPU_SYNC))
+		goto error_free;
+
+	return sgt;
+
+error_free:
+	sg_free_table(sgt);
+	kfree(sgt);
+	return ERR_PTR(-ENOMEM);
 }
 
 /**
- * amdgpu_gem_prime_res_obj - &drm_driver.gem_prime_res_obj implementation
- * @obj: GEM BO
+ * amdgpu_dma_buf_unmap - &dma_buf_ops.unmap_dma_buf implementation
+ * @attach: DMA-buf attachment
+ * @sgt: sg_table to unmap
+ * @dir: DMA direction
  *
- * Returns:
- * The BO's reservation object.
+ * This is called when a shared DMA buffer no longer needs to be accessible by
+ * another device. For now, simply unpins the buffer from GTT.
  */
-struct reservation_object *amdgpu_gem_prime_res_obj(struct drm_gem_object *obj)
+static void amdgpu_dma_buf_unmap(struct dma_buf_attachment *attach,
+				 struct sg_table *sgt,
+				 enum dma_data_direction dir)
 {
+	struct drm_gem_object *obj = attach->dmabuf->priv;
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
 
-	return bo->tbo.resv;
+	dma_unmap_sg(attach->dev, sgt->sgl, sgt->nents, dir);
+	sg_free_table(sgt);
+	kfree(sgt);
+	amdgpu_bo_unpin(bo);
 }
 
 /**
@@ -299,7 +304,7 @@ static int amdgpu_dma_buf_begin_cpu_access(struct dma_buf *dma_buf,
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(dma_buf->priv);
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
 	struct ttm_operation_ctx ctx = { true, false };
-	u32 domain = amdgpu_display_supported_domains(adev);
+	u32 domain = amdgpu_display_supported_domains(adev, bo->flags);
 	int ret;
 	bool reads = (direction == DMA_BIDIRECTIONAL ||
 		      direction == DMA_FROM_DEVICE);
@@ -322,10 +327,11 @@ static int amdgpu_dma_buf_begin_cpu_access(struct dma_buf *dma_buf,
 }
 
 const struct dma_buf_ops amdgpu_dmabuf_ops = {
-	.attach = amdgpu_dma_buf_map_attach,
-	.detach = amdgpu_dma_buf_map_detach,
-	.map_dma_buf = drm_gem_map_dma_buf,
-	.unmap_dma_buf = drm_gem_unmap_dma_buf,
+	.dynamic_mapping = true,
+	.attach = amdgpu_dma_buf_attach,
+	.detach = amdgpu_dma_buf_detach,
+	.map_dma_buf = amdgpu_dma_buf_map,
+	.unmap_dma_buf = amdgpu_dma_buf_unmap,
 	.release = drm_gem_dmabuf_release,
 	.begin_cpu_access = amdgpu_dma_buf_begin_cpu_access,
 	.mmap = drm_gem_dmabuf_mmap,
@@ -335,18 +341,15 @@ const struct dma_buf_ops amdgpu_dmabuf_ops = {
 
 /**
  * amdgpu_gem_prime_export - &drm_driver.gem_prime_export implementation
- * @dev: DRM device
  * @gobj: GEM BO
  * @flags: Flags such as DRM_CLOEXEC and DRM_RDWR.
  *
- * The main work is done by the &drm_gem_prime_export helper, which in turn
- * uses &amdgpu_gem_prime_res_obj.
+ * The main work is done by the &drm_gem_prime_export helper.
  *
  * Returns:
  * Shared DMA buffer representing the GEM BO from the given device.
  */
-struct dma_buf *amdgpu_gem_prime_export(struct drm_device *dev,
-					struct drm_gem_object *gobj,
+struct dma_buf *amdgpu_gem_prime_export(struct drm_gem_object *gobj,
 					int flags)
 {
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(gobj);
@@ -356,63 +359,56 @@ struct dma_buf *amdgpu_gem_prime_export(struct drm_device *dev,
 	    bo->flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID)
 		return ERR_PTR(-EPERM);
 
-	buf = drm_gem_prime_export(dev, gobj, flags);
-	if (!IS_ERR(buf)) {
-		buf->file->f_mapping = dev->anon_inode->i_mapping;
+	buf = drm_gem_prime_export(gobj, flags);
+	if (!IS_ERR(buf))
 		buf->ops = &amdgpu_dmabuf_ops;
-	}
 
 	return buf;
 }
 
 /**
- * amdgpu_gem_prime_import_sg_table - &drm_driver.gem_prime_import_sg_table
- * implementation
+ * amdgpu_dma_buf_create_obj - create BO for DMA-buf import
+ *
  * @dev: DRM device
- * @attach: DMA-buf attachment
- * @sg: Scatter/gather table
+ * @dma_buf: DMA-buf
  *
- * Imports shared DMA buffer memory exported by another device.
+ * Creates an empty SG BO for DMA-buf import.
  *
  * Returns:
  * A new GEM BO of the given DRM device, representing the memory
  * described by the given DMA-buf attachment and scatter/gather table.
  */
-struct drm_gem_object *
-amdgpu_gem_prime_import_sg_table(struct drm_device *dev,
-				 struct dma_buf_attachment *attach,
-				 struct sg_table *sg)
+static struct drm_gem_object *
+amdgpu_dma_buf_create_obj(struct drm_device *dev, struct dma_buf *dma_buf)
 {
-	struct reservation_object *resv = attach->dmabuf->resv;
+	struct dma_resv *resv = dma_buf->resv;
 	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_bo *bo;
 	struct amdgpu_bo_param bp;
 	int ret;
 
 	memset(&bp, 0, sizeof(bp));
-	bp.size = attach->dmabuf->size;
+	bp.size = dma_buf->size;
 	bp.byte_align = PAGE_SIZE;
 	bp.domain = AMDGPU_GEM_DOMAIN_CPU;
 	bp.flags = 0;
 	bp.type = ttm_bo_type_sg;
 	bp.resv = resv;
-	ww_mutex_lock(&resv->lock, NULL);
+	dma_resv_lock(resv, NULL);
 	ret = amdgpu_bo_create(adev, &bp, &bo);
 	if (ret)
 		goto error;
 
-	bo->tbo.sg = sg;
-	bo->tbo.ttm->sg = sg;
 	bo->allowed_domains = AMDGPU_GEM_DOMAIN_GTT;
 	bo->preferred_domains = AMDGPU_GEM_DOMAIN_GTT;
-	if (attach->dmabuf->ops != &amdgpu_dmabuf_ops)
+	if (dma_buf->ops != &amdgpu_dmabuf_ops)
 		bo->prime_shared_count = 1;
 
-	ww_mutex_unlock(&resv->lock);
-	return &bo->gem_base;
+	dma_resv_unlock(resv);
+	return &bo->tbo.base;
 
 error:
-	ww_mutex_unlock(&resv->lock);
+	dma_resv_unlock(resv);
 	return ERR_PTR(ret);
 }
 
@@ -421,15 +417,15 @@ error:
  * @dev: DRM device
  * @dma_buf: Shared DMA buffer
  *
- * The main work is done by the &drm_gem_prime_import helper, which in turn
- * uses &amdgpu_gem_prime_import_sg_table.
+ * Import a dma_buf into a the driver and potentially create a new GEM object.
  *
  * Returns:
  * GEM BO representing the shared DMA buffer for the given device.
  */
 struct drm_gem_object *amdgpu_gem_prime_import(struct drm_device *dev,
-					    struct dma_buf *dma_buf)
+					       struct dma_buf *dma_buf)
 {
+	struct dma_buf_attachment *attach;
 	struct drm_gem_object *obj;
 
 	if (dma_buf->ops == &amdgpu_dmabuf_ops) {
@@ -444,5 +440,17 @@ struct drm_gem_object *amdgpu_gem_prime_import(struct drm_device *dev,
 		}
 	}
 
-	return drm_gem_prime_import(dev, dma_buf);
+	obj = amdgpu_dma_buf_create_obj(dev, dma_buf);
+	if (IS_ERR(obj))
+		return obj;
+
+	attach = dma_buf_dynamic_attach(dma_buf, dev->dev, true);
+	if (IS_ERR(attach)) {
+		drm_gem_object_put(obj);
+		return ERR_CAST(attach);
+	}
+
+	get_dma_buf(dma_buf);
+	obj->import_attach = attach;
+	return obj;
 }