lk/linux-kernel-test
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

drm/i915: Only enforce fence limits inside the GTT.

Browse Source 
So long as we adhere to the fence registers rules for alignment and no
overlaps (including with unfenced accesses to linear memory) and account
for the tiled access in our size allocation, we do not have to allocate
the full fenced region for the object. This allows us to fight the bloat
tiling imposed on pre-i965 chipsets and frees up RAM for real use. [Inside
the GTT we still suffer the additional alignment constraints, so it doesn't
magic allow us to render larger scenes without stalls -- we need the
expanded GTT and fence pipelining to overcome those...]

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
This commit is contained in:
Chris Wilson
2010-09-24 21:15:47 +01:00
parent 7465378fd7
commit a00b10c360
9 changed files with 197 additions and 155 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
drivers/gpu/drm/i915/intel_ringbuffer.c
View File

@@ -547,7 +547,7 @@ static int init_status_page(struct intel_ring_buffer *ring)
obj_priv = to_intel_bo(obj);
obj_priv->agp_type = AGP_USER_CACHED_MEMORY;
ret = i915_gem_object_pin(obj, 4096, true);
ret = i915_gem_object_pin(obj, 4096, true, false);
if (ret != 0) {
goto err_unref;
}
@@ -603,7 +603,7 @@ int intel_init_ring_buffer(struct drm_device *dev,
ring->gem_object = obj;
ret = i915_gem_object_pin(obj, PAGE_SIZE, true);
ret = i915_gem_object_pin(obj, PAGE_SIZE, true, false);
if (ret)
goto err_unref;