libvpx: Pull from upstream

source/libvpx/vp9/encoder/x86/vp9_avg_intrin_sse2.c

 /*
  *  Copyright (c) 2014 The WebM project authors. All Rights Reserved.
  *
  *  Use of this source code is governed by a BSD-style license
  *  that can be found in the LICENSE file in the root of the source
  *  tree. An additional intellectual property rights grant can be found
  *  in the file PATENTS.  All contributing project authors may
  *  be found in the AUTHORS file in the root of the source tree.
  */
 
@@ skipping 94 matching lines @@
   hbuf += 8;
   _mm_storeu_si128((__m128i *)hbuf, s1);
 }
 
 int16_t vp9_int_pro_col_sse2(uint8_t const *ref, const int width) {
   __m128i zero = _mm_setzero_si128();
   __m128i src_line = _mm_load_si128((const __m128i *)ref);
   __m128i s0 = _mm_sad_epu8(src_line, zero);
   __m128i s1;
   int i;
-  const int norm_factor = 3 + (width >> 5);
 
   for (i = 16; i < width; i += 16) {
     ref += 16;
     src_line = _mm_load_si128((const __m128i *)ref);
     s1 = _mm_sad_epu8(src_line, zero);
     s0 = _mm_adds_epu16(s0, s1);
   }
 
   s1 = _mm_srli_si128(s0, 8);
   s0 = _mm_adds_epu16(s0, s1);
 
-  return _mm_extract_epi16(s0, 0) >> norm_factor;
+  return _mm_extract_epi16(s0, 0);
 }
 
 int vp9_vector_var_sse2(int16_t const *ref, int16_t const *src,
                         const int bwl) {
   int idx;
   int width = 4 << bwl;
   int16_t mean;
   __m128i v0 = _mm_loadu_si128((const __m128i *)ref);
   __m128i v1 = _mm_load_si128((const __m128i *)src);
   __m128i diff = _mm_subs_epi16(v0, v1);
@@ skipping 25 matching lines @@
 
   v1  = _mm_srli_si128(sse, 8);
   sse = _mm_add_epi32(sse, v1);
   v1  = _mm_srli_epi64(sse, 32);
   sse = _mm_add_epi32(sse, v1);
 
   mean = _mm_extract_epi16(sum, 0);
 
   return _mm_cvtsi128_si32(sse) - ((mean * mean) >> (bwl + 2));
 }