Each importer can now provide an invalidate_mappings callback.

This allows the exporter to provide the mappings without the need to pin the backing store.

v2: don't try to invalidate mappings when the callback is NULL, lock the reservation obj while using the attachments, add helper to set the callback v3: move flag for invalidation support into the DMA-buf, use new attach_info structure to set the callback v4: use importer_priv field instead of mangling exporter priv.

Signed-off-by: Christian König christian.koenig@amd.com --- drivers/dma-buf/dma-buf.c | 44 ++++++++++++++++++++++++++++++++++++++++++++ include/linux/dma-buf.h | 36 ++++++++++++++++++++++++++++++++++-- 2 files changed, 78 insertions(+), 2 deletions(-)

diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c index 4b46982c6d9c..ffdaab10e2f2 100644 --- a/drivers/dma-buf/dma-buf.c +++ b/drivers/dma-buf/dma-buf.c @@ -565,6 +565,8 @@ struct dma_buf_attachment *dma_buf_attach(const struct dma_buf_attach_info *info

attach->dev = info->dev; attach->dmabuf = dmabuf; + attach->importer_priv = info->importer_priv; + attach->invalidate = info->invalidate;

mutex_lock(&dmabuf->lock);

@@ -573,7 +575,9 @@ struct dma_buf_attachment *dma_buf_attach(const struct dma_buf_attach_info *info if (ret) goto err_attach; } + reservation_object_lock(dmabuf->resv, NULL); list_add(&attach->node, &dmabuf->attachments); + reservation_object_unlock(dmabuf->resv);

mutex_unlock(&dmabuf->lock); return attach; @@ -599,7 +603,9 @@ void dma_buf_detach(struct dma_buf *dmabuf, struct dma_buf_attachment *attach) return;

mutex_lock(&dmabuf->lock); + reservation_object_lock(dmabuf->resv, NULL); list_del(&attach->node); + reservation_object_unlock(dmabuf->resv); if (dmabuf->ops->detach) dmabuf->ops->detach(dmabuf, attach);

@@ -633,10 +639,23 @@ struct sg_table *dma_buf_map_attachment(struct dma_buf_attachment *attach, if (WARN_ON(!attach || !attach->dmabuf)) return ERR_PTR(-EINVAL);

+ /* + * Mapping a DMA-buf can trigger its invalidation, prevent sending this + * event to the caller by temporary removing this attachment from the + * list. + */ + if (attach->invalidate) { + reservation_object_assert_held(attach->dmabuf->resv); + list_del(&attach->node); + } + sg_table = attach->dmabuf->ops->map_dma_buf(attach, direction); if (!sg_table) sg_table = ERR_PTR(-ENOMEM);

+ if (attach->invalidate) + list_add(&attach->node, &attach->dmabuf->attachments); + return sg_table; } EXPORT_SYMBOL_GPL(dma_buf_map_attachment); @@ -657,6 +676,9 @@ void dma_buf_unmap_attachment(struct dma_buf_attachment *attach, { might_sleep();

+ if (attach->invalidate) + reservation_object_assert_held(attach->dmabuf->resv); + if (WARN_ON(!attach || !attach->dmabuf || !sg_table)) return;

@@ -665,6 +687,26 @@ void dma_buf_unmap_attachment(struct dma_buf_attachment *attach, } EXPORT_SYMBOL_GPL(dma_buf_unmap_attachment);

+/** + * dma_buf_invalidate_mappings - invalidate all mappings of this dma_buf + * + * @dmabuf: [in] buffer which mappings should be invalidated + * + * Informs all attachmenst that they need to destroy and recreated all their + * mappings. + */ +void dma_buf_invalidate_mappings(struct dma_buf *dmabuf) +{ + struct dma_buf_attachment *attach; + + reservation_object_assert_held(dmabuf->resv); + + list_for_each_entry(attach, &dmabuf->attachments, node) + if (attach->invalidate) + attach->invalidate(attach); +} +EXPORT_SYMBOL_GPL(dma_buf_invalidate_mappings); + /** * DOC: cpu access * @@ -1122,10 +1164,12 @@ static int dma_buf_debug_show(struct seq_file *s, void *unused) seq_puts(s, "\tAttached Devices:\n"); attach_count = 0;

+ reservation_object_lock(buf_obj->resv, NULL); list_for_each_entry(attach_obj, &buf_obj->attachments, node) { seq_printf(s, "\t%s\n", dev_name(attach_obj->dev)); attach_count++; } + reservation_object_unlock(buf_obj->resv);

seq_printf(s, "Total %d devices attached\n\n", attach_count); diff --git a/include/linux/dma-buf.h b/include/linux/dma-buf.h index 414b4dde5eb7..566503dd2b4f 100644 --- a/include/linux/dma-buf.h +++ b/include/linux/dma-buf.h @@ -270,6 +270,8 @@ struct dma_buf_ops { * @poll: for userspace poll support * @cb_excl: for userspace poll support * @cb_shared: for userspace poll support + * @invalidation_supported: True when the exporter supports unpinned operation + * using the reservation lock. * * This represents a shared buffer, created by calling dma_buf_export(). The * userspace representation is a normal file descriptor, which can be created by @@ -293,6 +295,7 @@ struct dma_buf { struct list_head list_node; void *priv; struct reservation_object *resv; + bool invalidation_supported;

/* poll support */ wait_queue_head_t poll; @@ -311,6 +314,7 @@ struct dma_buf { * @dev: device attached to the buffer. * @node: list of dma_buf_attachment. * @priv: exporter specific attachment data. + * @importer_priv: importer specific attachment data. * * This structure holds the attachment information between the dma_buf buffer * and its user device(s). The list contains one attachment struct per device @@ -326,6 +330,29 @@ struct dma_buf_attachment { struct device *dev; struct list_head node; void *priv; + void *importer_priv; + + /** + * @invalidate: + * + * Optional callback provided by the importer of the dma-buf. + * + * If provided the exporter can avoid pinning the backing store while + * mappings exists. + * + * The function is called with the lock of the reservation object + * associated with the dma_buf held and the mapping function must be + * called with this lock held as well. This makes sure that no mapping + * is created concurrently with an ongoing invalidation. + * + * After the callback all existing mappings are still valid until all + * fences in the dma_bufs reservation object are signaled, but should be + * destroyed by the importer as soon as possible. + * + * New mappings can be created immediately, but can't be used before the + * exclusive fence in the dma_bufs reservation object is signaled. + */ + void (*invalidate)(struct dma_buf_attachment *attach); };

/** @@ -364,8 +391,10 @@ struct dma_buf_export_info {

/** * struct dma_buf_attach_info - holds information needed to attach to a dma_buf - * @dmabuf: the exported dma_buf - * @dev: the device which wants to import the attachment + * @dmabuf: the exported dma_buf + * @dev: the device which wants to import the attachment + * @importer_priv: private data of importer to this attachment + * @invalidate: callback to use for invalidating mappings * * This structure holds the information required to attach to a buffer. Used * with dma_buf_attach() only. @@ -373,6 +402,8 @@ struct dma_buf_export_info { struct dma_buf_attach_info { struct dma_buf *dmabuf; struct device *dev; + void *importer_priv; + void (*invalidate)(struct dma_buf_attachment *attach); };

/** @@ -404,6 +435,7 @@ struct sg_table *dma_buf_map_attachment(struct dma_buf_attachment *, enum dma_data_direction); void dma_buf_unmap_attachment(struct dma_buf_attachment *, struct sg_table *, enum dma_data_direction); +void dma_buf_invalidate_mappings(struct dma_buf *dma_buf); int dma_buf_begin_cpu_access(struct dma_buf *dma_buf, enum dma_data_direction dir); int dma_buf_end_cpu_access(struct dma_buf *dma_buf,

Instead of relying on the DRM functions just implement our own export functions. This adds support for taking care of unpinned DMA-buf.

v2: fix unintended recursion, remove debugging leftovers

Signed-off-by: Christian König christian.koenig@amd.com --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 1 - drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 1 - drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 5 ++ drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c | 108 ++++++++++++++++++----------- 4 files changed, 72 insertions(+), 43 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 2babfad1fd7f..3e1727251edc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -371,7 +371,6 @@ int amdgpu_gem_object_open(struct drm_gem_object *obj, void amdgpu_gem_object_close(struct drm_gem_object *obj, struct drm_file *file_priv); unsigned long amdgpu_gem_timeout(uint64_t timeout_ns); -struct sg_table *amdgpu_gem_prime_get_sg_table(struct drm_gem_object *obj); struct drm_gem_object * amdgpu_gem_prime_import_sg_table(struct drm_device *dev, struct dma_buf_attachment *attach, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 0b19482b36b8..6107b845d8a8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -890,7 +890,6 @@ static struct drm_driver kms_driver = { .gem_prime_export = amdgpu_gem_prime_export, .gem_prime_import = amdgpu_gem_prime_import, .gem_prime_res_obj = amdgpu_gem_prime_res_obj, - .gem_prime_get_sg_table = amdgpu_gem_prime_get_sg_table, .gem_prime_import_sg_table = amdgpu_gem_prime_import_sg_table, .gem_prime_vmap = amdgpu_gem_prime_vmap, .gem_prime_vunmap = amdgpu_gem_prime_vunmap, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 9e23d6f6f3f3..15bb48cebbc4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -31,6 +31,7 @@ */ #include <linux/list.h> #include <linux/slab.h> +#include <linux/dma-buf.h> #include <drm/drmP.h> #include <drm/amdgpu_drm.h> #include <drm/drm_cache.h> @@ -953,6 +954,10 @@ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo,

amdgpu_bo_kunmap(abo);

+ if (abo->gem_base.dma_buf && !abo->gem_base.import_attach && + bo->mem.mem_type != TTM_PL_SYSTEM) + dma_buf_invalidate_mappings(abo->gem_base.dma_buf); + /* remember the eviction */ if (evict) atomic64_inc(&adev->num_evictions); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c index 4b584cb75bf4..7fef95f0fed1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c @@ -32,14 +32,6 @@

static const struct dma_buf_ops amdgpu_dmabuf_ops;

-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); -} - void *amdgpu_gem_prime_vmap(struct drm_gem_object *obj) { struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj); @@ -126,23 +118,22 @@ amdgpu_gem_prime_import_sg_table(struct drm_device *dev, return ERR_PTR(ret); }

-static int amdgpu_gem_map_attach(struct dma_buf *dma_buf, - struct device *target_dev, - struct dma_buf_attachment *attach) +static struct sg_table * +amdgpu_gem_map_dma_buf(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); + struct sg_table *sgt; long r;

- r = drm_gem_map_attach(dma_buf, target_dev, attach); - if (r) - return r; - - r = amdgpu_bo_reserve(bo, false); - if (unlikely(r != 0)) - goto error_detach; - + if (!attach->invalidate) { + r = amdgpu_bo_reserve(bo, false); + if (unlikely(r != 0)) + return ERR_PTR(r); + }

if (attach->dev->driver != adev->dev->driver) { /* @@ -158,42 +149,77 @@ static int amdgpu_gem_map_attach(struct dma_buf *dma_buf, } }

- /* pin buffer into GTT */ - r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT, NULL); - if (r) + if (!attach->invalidate) { + /* pin buffer into GTT */ + r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT, NULL); + if (r) + goto error_unreserve; + + } else { + /* move buffer into GTT */ + struct ttm_operation_ctx ctx = { false, false }; + + amdgpu_ttm_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_GTT); + r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx); + if (r) + goto error_unreserve; + } + + sgt = drm_prime_pages_to_sg(bo->tbo.ttm->pages, bo->tbo.num_pages); + if (IS_ERR(sgt)) { + r = PTR_ERR(sgt); goto error_unreserve; + } + + if (!dma_map_sg_attrs(attach->dev, sgt->sgl, sgt->nents, dir, + DMA_ATTR_SKIP_CPU_SYNC)) + goto error_free;

if (attach->dev->driver != adev->dev->driver) bo->prime_shared_count++;

+ if (!attach->invalidate) + amdgpu_bo_unreserve(bo); + + return sgt; + +error_free: + sg_free_table(sgt); + kfree(sgt); + r = -ENOMEM; + error_unreserve: - amdgpu_bo_unreserve(bo); + if (!attach->invalidate) + amdgpu_bo_unreserve(bo);

-error_detach: - if (r) - drm_gem_map_detach(dma_buf, attach); - return r; + return ERR_PTR(r); }

-static void amdgpu_gem_map_detach(struct dma_buf *dma_buf, - struct dma_buf_attachment *attach) +static void amdgpu_gem_unmap_dma_buf(struct dma_buf_attachment *attach, + struct sg_table *sgt, + 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;

- ret = amdgpu_bo_reserve(bo, true); - if (unlikely(ret != 0)) - goto error; + if (!attach->invalidate) { + amdgpu_bo_reserve(bo, true); + amdgpu_bo_unpin(bo); + }

- amdgpu_bo_unpin(bo); if (attach->dev->driver != adev->dev->driver && bo->prime_shared_count) bo->prime_shared_count--; - amdgpu_bo_unreserve(bo);

-error: - drm_gem_map_detach(dma_buf, attach); + if (!attach->invalidate) + amdgpu_bo_unreserve(bo); + + if (sgt) { + dma_unmap_sg(attach->dev, sgt->sgl, sgt->nents, dir); + sg_free_table(sgt); + kfree(sgt); + } }

struct reservation_object *amdgpu_gem_prime_res_obj(struct drm_gem_object *obj) @@ -232,10 +258,8 @@ static int amdgpu_gem_begin_cpu_access(struct dma_buf *dma_buf, }

static const struct dma_buf_ops amdgpu_dmabuf_ops = { - .attach = amdgpu_gem_map_attach, - .detach = amdgpu_gem_map_detach, - .map_dma_buf = drm_gem_map_dma_buf, - .unmap_dma_buf = drm_gem_unmap_dma_buf, + .map_dma_buf = amdgpu_gem_map_dma_buf, + .unmap_dma_buf = amdgpu_gem_unmap_dma_buf, .release = drm_gem_dmabuf_release, .begin_cpu_access = amdgpu_gem_begin_cpu_access, .map = drm_gem_dmabuf_kmap, @@ -260,6 +284,8 @@ struct dma_buf *amdgpu_gem_prime_export(struct drm_device *dev,

buf = drm_gem_prime_export(dev, gobj, flags); if (!IS_ERR(buf)) { + buf->invalidation_supported = + !!(bo->preferred_domains & AMDGPU_GEM_DOMAIN_GTT); buf->file->f_mapping = dev->anon_inode->i_mapping; buf->ops = &amdgpu_dmabuf_ops; }