[PATCH 22/22] drm/i915: only clear partial pt upon allocation

[Matthew Auld]

If we are going to fill the entire page anyway when inserting PTEs then
there's not much point in clearing it before hand. For particularly
large objects the fill_page_dma tends to rank pretty highly in perf when
running benchmarks like gem_exec_fault, for both the non-huge and huge
cases.

Signed-off-by: Matthew Auld <matthew.auld at intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen at linux.intel.com>
Cc: Chris Wilson <chris at chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_gem_gtt.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index d643fc486451..529547d35286 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -1230,7 +1230,8 @@ static int gen8_ppgtt_alloc_pd(struct i915_address_space *vm,
 			if (IS_ERR(pt))
 				goto unwind;
 
-			gen8_initialize_pt(vm, pt);
+			if (length - start < SZ_2M)
+				gen8_initialize_pt(vm, pt);
 
 			gen8_ppgtt_set_pde(vm, pd, pt, pde);
 			pd->used_pdes++;
-- 
2.9.4