diff options
author | Chris Wilson <chris@chris-wilson.co.uk> | 2016-10-28 13:58:44 +0100 |
---|---|---|
committer | Chris Wilson <chris@chris-wilson.co.uk> | 2016-10-28 20:53:50 +0100 |
commit | d07f0e59b2c762584478920cd2d11fba2980a94a (patch) | |
tree | a9fc0cade87570bdec356493e6744ee5223bdaaf /drivers/gpu/drm/i915/i915_gem_dmabuf.c | |
parent | f0cd518206e1a47e57bc251e1faba9d38eadcc59 (diff) | |
download | op-kernel-dev-d07f0e59b2c762584478920cd2d11fba2980a94a.zip op-kernel-dev-d07f0e59b2c762584478920cd2d11fba2980a94a.tar.gz |
drm/i915: Move GEM activity tracking into a common struct reservation_object
In preparation to support many distinct timelines, we need to expand the
activity tracking on the GEM object to handle more than just a request
per engine. We already use the struct reservation_object on the dma-buf
to handle many fence contexts, so integrating that into the GEM object
itself is the preferred solution. (For example, we can now share the same
reservation_object between every consumer/producer using this buffer and
skip the manual import/export via dma-buf.)
v2: Reimplement busy-ioctl (by walking the reservation object), postpone
the ABI change for another day. Similarly use the reservation object to
find the last_write request (if active and from i915) for choosing
display CS flips.
Caveats:
* busy-ioctl: busy-ioctl only reports on the native fences, it will not
warn of stalls (in set-domain-ioctl, pread/pwrite etc) if the object is
being rendered to by external fences. It also will not report the same
busy state as wait-ioctl (or polling on the dma-buf) in the same
circumstances. On the plus side, it does retain reporting of which
*i915* engines are engaged with this object.
* non-blocking atomic modesets take a step backwards as the wait for
render completion blocks the ioctl. This is fixed in a subsequent
patch to use a fence instead for awaiting on the rendering, see
"drm/i915: Restore nonblocking awaits for modesetting"
* dynamic array manipulation for shared-fences in reservation is slower
than the previous lockless static assignment (e.g. gem_exec_lut_handle
runtime on ivb goes from 42s to 66s), mainly due to atomic operations
(maintaining the fence refcounts).
* loss of object-level retirement callbacks, emulated by VMA retirement
tracking.
* minor loss of object-level last activity information from debugfs,
could be replaced with per-vma information if desired
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20161028125858.23563-21-chris@chris-wilson.co.uk
Diffstat (limited to 'drivers/gpu/drm/i915/i915_gem_dmabuf.c')
-rw-r--r-- | drivers/gpu/drm/i915/i915_gem_dmabuf.c | 53 |
1 files changed, 3 insertions, 50 deletions
diff --git a/drivers/gpu/drm/i915/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/i915_gem_dmabuf.c index 4d45f20..5e38299 100644 --- a/drivers/gpu/drm/i915/i915_gem_dmabuf.c +++ b/drivers/gpu/drm/i915/i915_gem_dmabuf.c @@ -211,60 +211,17 @@ static const struct dma_buf_ops i915_dmabuf_ops = { .end_cpu_access = i915_gem_end_cpu_access, }; -static void export_fences(struct drm_i915_gem_object *obj, - struct dma_buf *dma_buf) -{ - struct reservation_object *resv = dma_buf->resv; - struct drm_i915_gem_request *req; - unsigned long active; - int idx; - - active = __I915_BO_ACTIVE(obj); - if (!active) - return; - - /* Serialise with execbuf to prevent concurrent fence-loops */ - mutex_lock(&obj->base.dev->struct_mutex); - - /* Mark the object for future fences before racily adding old fences */ - obj->base.dma_buf = dma_buf; - - ww_mutex_lock(&resv->lock, NULL); - - for_each_active(active, idx) { - req = i915_gem_active_get(&obj->last_read[idx], - &obj->base.dev->struct_mutex); - if (!req) - continue; - - if (reservation_object_reserve_shared(resv) == 0) - reservation_object_add_shared_fence(resv, &req->fence); - - i915_gem_request_put(req); - } - - req = i915_gem_active_get(&obj->last_write, - &obj->base.dev->struct_mutex); - if (req) { - reservation_object_add_excl_fence(resv, &req->fence); - i915_gem_request_put(req); - } - - ww_mutex_unlock(&resv->lock); - mutex_unlock(&obj->base.dev->struct_mutex); -} - struct dma_buf *i915_gem_prime_export(struct drm_device *dev, struct drm_gem_object *gem_obj, int flags) { struct drm_i915_gem_object *obj = to_intel_bo(gem_obj); DEFINE_DMA_BUF_EXPORT_INFO(exp_info); - struct dma_buf *dma_buf; exp_info.ops = &i915_dmabuf_ops; exp_info.size = gem_obj->size; exp_info.flags = flags; exp_info.priv = gem_obj; + exp_info.resv = obj->resv; if (obj->ops->dmabuf_export) { int ret = obj->ops->dmabuf_export(obj); @@ -272,12 +229,7 @@ struct dma_buf *i915_gem_prime_export(struct drm_device *dev, return ERR_PTR(ret); } - dma_buf = drm_gem_dmabuf_export(dev, &exp_info); - if (IS_ERR(dma_buf)) - return dma_buf; - - export_fences(obj, dma_buf); - return dma_buf; + return drm_gem_dmabuf_export(dev, &exp_info); } static struct sg_table * @@ -335,6 +287,7 @@ struct drm_gem_object *i915_gem_prime_import(struct drm_device *dev, drm_gem_private_object_init(dev, &obj->base, dma_buf->size); i915_gem_object_init(obj, &i915_gem_object_dmabuf_ops); obj->base.import_attach = attach; + obj->resv = dma_buf->resv; /* We use GTT as shorthand for a coherent domain, one that is * neither in the GPU cache nor in the CPU cache, where all |