From 4f1836453e42edcc7bb3c53c6c38534697ca77b0 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Mon, 4 Mar 2019 23:06:46 +0000 Subject: [PATCH] drm/i915/gtt: Use optimised memset32/64 for clearing PTE Replace the open-coded memset loops with the memset32/64 routines that reduce to a single instruction or two: add/remove: 0/0 grow/shrink: 0/2 up/down: 0/-83 (-83) Function old new delta gen6_ppgtt_clear_range 371 344 -27 gen8_ppgtt_clear_pd 575 519 -56 Signed-off-by: Chris Wilson Cc: Matthew Auld Cc: Mika Kuoppala Reviewed-by: Matthew Auld Link: https://patchwork.freedesktop.org/patch/msgid/20190304230646.23714-1-chris@chris-wilson.co.uk --- drivers/gpu/drm/i915/i915_gem_gtt.c | 12 +++--------- 1 file changed, 3 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index 7e79691664e5..f97cc7b437f2 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -809,8 +809,6 @@ static bool gen8_ppgtt_clear_pt(const struct i915_address_space *vm, u64 start, u64 length) { unsigned int num_entries = gen8_pte_count(start, length); - unsigned int pte = gen8_pte_index(start); - unsigned int pte_end = pte + num_entries; gen8_pte_t *vaddr; GEM_BUG_ON(num_entries > pt->used_ptes); @@ -820,8 +818,7 @@ static bool gen8_ppgtt_clear_pt(const struct i915_address_space *vm, return true; vaddr = kmap_atomic_px(pt); - while (pte < pte_end) - vaddr[pte++] = vm->scratch_pte; + memset64(vaddr + gen8_pte_index(start), vm->scratch_pte, num_entries); kunmap_atomic(vaddr); return false; @@ -1672,8 +1669,7 @@ static void gen6_ppgtt_clear_range(struct i915_address_space *vm, while (num_entries) { struct i915_page_table *pt = ppgtt->base.pd.page_table[pde++]; - const unsigned int end = min(pte + num_entries, GEN6_PTES); - const unsigned int count = end - pte; + const unsigned int count = min(num_entries, GEN6_PTES - pte); gen6_pte_t *vaddr; GEM_BUG_ON(pt == vm->scratch_pt); @@ -1693,9 +1689,7 @@ static void gen6_ppgtt_clear_range(struct i915_address_space *vm, */ vaddr = kmap_atomic_px(pt); - do { - vaddr[pte++] = scratch_pte; - } while (pte < end); + memset32(vaddr + pte, scratch_pte, count); kunmap_atomic(vaddr); pte = 0; -- 2.45.2