libvpx: Pull from upstream

source/libvpx/vp9/encoder/x86/vp9_dct32x32_avx2.c

 /*
  *  Copyright (c) 2012 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.
  */
 
 #include <immintrin.h>  // AVX2
 #include "vp9/common/vp9_idct.h"  // for cospi constants
 #include "vpx_ports/mem.h"
 
 #define pair256_set_epi16(a, b) \
-  _mm256_set_epi16(b, a, b, a, b, a, b, a, b, a, b, a, b, a, b, a)
+  _mm256_set_epi16((int16_t)(b), (int16_t)(a), (int16_t)(b), (int16_t)(a), \
+                   (int16_t)(b), (int16_t)(a), (int16_t)(b), (int16_t)(a), \
+                   (int16_t)(b), (int16_t)(a), (int16_t)(b), (int16_t)(a), \
+                   (int16_t)(b), (int16_t)(a), (int16_t)(b), (int16_t)(a))
 
 #define pair256_set_epi32(a, b) \
-  _mm256_set_epi32(b, a, b, a, b, a, b, a)
-
-
-
+  _mm256_set_epi32((int)(b), (int)(a), (int)(b), (int)(a), \
+                   (int)(b), (int)(a), (int)(b), (int)(a))
 
 #if FDCT32x32_HIGH_PRECISION
 static INLINE __m256i k_madd_epi32_avx2(__m256i a, __m256i b) {
   __m256i buf0, buf1;
   buf0 = _mm256_mul_epu32(a, b);
   a = _mm256_srli_epi64(a, 32);
   b = _mm256_srli_epi64(b, 32);
   buf1 = _mm256_mul_epu32(a, b);
   return _mm256_add_epi64(buf0, buf1);
 }
@@ skipping 10 matching lines @@
   // Calculate pre-multiplied strides
   const int str1 = stride;
   const int str2 = 2 * stride;
   const int str3 = 2 * stride + str1;
   // We need an intermediate buffer between passes.
   DECLARE_ALIGNED(32, int16_t, intermediate[32 * 32]);
   // Constants
   //    When we use them, in one case, they are all the same. In all others
   //    it's a pair of them that we need to repeat four times. This is done
   //    by constructing the 32 bit constant corresponding to that pair.
-  const __m256i k__cospi_p16_p16 = _mm256_set1_epi16(+cospi_16_64);
+  const __m256i k__cospi_p16_p16 = _mm256_set1_epi16((int16_t)cospi_16_64);
   const __m256i k__cospi_p16_m16 = pair256_set_epi16(+cospi_16_64, -cospi_16_64);
   const __m256i k__cospi_m08_p24 = pair256_set_epi16(-cospi_8_64,   cospi_24_64);
   const __m256i k__cospi_m24_m08 = pair256_set_epi16(-cospi_24_64, -cospi_8_64);
   const __m256i k__cospi_p24_p08 = pair256_set_epi16(+cospi_24_64,  cospi_8_64);
   const __m256i k__cospi_p12_p20 = pair256_set_epi16(+cospi_12_64,  cospi_20_64);
   const __m256i k__cospi_m20_p12 = pair256_set_epi16(-cospi_20_64,  cospi_12_64);
   const __m256i k__cospi_m04_p28 = pair256_set_epi16(-cospi_4_64,   cospi_28_64);
   const __m256i k__cospi_p28_p04 = pair256_set_epi16(+cospi_28_64,  cospi_4_64);
   const __m256i k__cospi_m28_m04 = pair256_set_epi16(-cospi_28_64, -cospi_4_64);
   const __m256i k__cospi_m12_m20 = pair256_set_epi16(-cospi_12_64, -cospi_20_64);
@@ skipping 2637 matching lines @@
           _mm_storeu_si128((__m128i *)(output_nextStep + 6 * 32), _mm256_extractf128_si256(tr2_6,1));
           _mm_storeu_si128((__m128i *)(output_nextStep + 7 * 32), _mm256_extractf128_si256(tr2_7,1));
           // Process next 8x8
           output_currStep += 8;
           output_nextStep += 8;
         }
       }
     }
   }
 }  // NOLINT