Commit c7c6e46f authored Aug 16, 2017 by Chris Wilson

drm/i915: Convert execbuf to use struct-of-array packing for critical fields

When userspace is doing most of the work, avoiding relocs (using
NO_RELOC) and opting out of implicit synchronisation (using ASYNC), we
still spend a lot of time processing the arrays in execbuf, even though
we now should have nothing to do most of the time. One issue that
becomes readily apparent in profiling anv is that iterating over the
large execobj[] is unfriendly to the loop prefetchers of the CPU and it
much prefers iterating over a pair of arrays rather than one big array.

v2: Clear vma[] on construction to handle errors during vma lookup

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170816085210.4199-3-chris@chris-wilson.co.uk

parent 8bcbfb12

drivers/gpu/drm/i915/i915_gem_evict.c

+2 −2

Original line number	Diff line number	Diff line
		@@ -318,8 +318,8 @@ int i915_gem_evict_for_node(struct i915_address_space *vm,
		/* Overlap of objects in the same batch? */
		if (i915_vma_is_pinned(vma)) {
		ret = -ENOSPC;
		if (vma->exec_entry &&
		vma->exec_entry->flags & EXEC_OBJECT_PINNED)
		if (vma->exec_flags &&
		*vma->exec_flags & EXEC_OBJECT_PINNED)
		ret = -EINVAL;
		break;
		}

drivers/gpu/drm/i915/i915_gem_execbuffer.c

+153 −148

File changed.

Preview size limit exceeded, changes collapsed.

drivers/gpu/drm/i915/i915_vma.h

+1 −1

Original line number	Diff line number	Diff line
		@@ -112,7 +112,7 @@ struct i915_vma {
		/**
		* Used for performing relocations during execbuffer insertion.
		*/
		struct drm_i915_gem_exec_object2 *exec_entry;
		unsigned int *exec_flags;
		struct hlist_node exec_node;
		u32 exec_handle;

Admin message