diff --git a/src/strategies/avx2/avx2_common_functions.h b/src/strategies/avx2/avx2_common_functions.h
index 3492abf1..3b6063ad 100644
--- a/src/strategies/avx2/avx2_common_functions.h
+++ b/src/strategies/avx2/avx2_common_functions.h
@@ -123,6 +123,7 @@ static INLINE void get_first_last_nz_int16(__m256i ints, int32_t *first, int32_t
   *last = (31 - (int32_t)_lzcnt_u32(nonzero_bytes)) >> 1;
 }
 
+/* MOVED TO SAO-AVX2.C WHERE THIS IS USED
 int32_t FIX_W32 kvz_hsum_8x32b(const __m256i v)
 {
   __m256i sum1 = v;
@@ -137,5 +138,5 @@ int32_t FIX_W32 kvz_hsum_8x32b(const __m256i v)
   int32_t sum9 = _mm_cvtsi128_si32       (sum8);
   return  sum9;
 }
-
+*/
 #endif
diff --git a/src/strategies/avx2/sao-avx2.c b/src/strategies/avx2/sao-avx2.c
index 5dc1a4da..5bd2dd91 100644
--- a/src/strategies/avx2/sao-avx2.c
+++ b/src/strategies/avx2/sao-avx2.c
@@ -271,6 +271,21 @@ static INLINE __m256i FIX_W32 do_one_edge_ymm(const __m256i a,
   return calc_diff_off_delta(diff_lo, diff_hi, offset, orig);
 }
 
+int32_t FIX_W32 kvz_hsum_8x32b(const __m256i v)
+{
+  __m256i sum1 = v;
+  __m256i sum2 = _mm256_permute4x64_epi64(sum1, _MM_SHUFFLE(1, 0, 3, 2));
+  __m256i sum3 = _mm256_add_epi32(sum1, sum2);
+  __m256i sum4 = _mm256_shuffle_epi32(sum3, _MM_SHUFFLE(1, 0, 3, 2));
+  __m256i sum5 = _mm256_add_epi32(sum3, sum4);
+  __m256i sum6 = _mm256_shuffle_epi32(sum5, _MM_SHUFFLE(2, 3, 0, 1));
+  __m256i sum7 = _mm256_add_epi32(sum5, sum6);
+
+  __m128i sum8 = _mm256_castsi256_si128(sum7);
+  int32_t sum9 = _mm_cvtsi128_si32(sum8);
+  return  sum9;
+}
+
 static int32_t sao_edge_ddistortion_avx2(const kvz_pixel *orig_data,
                                          const kvz_pixel *rec_data,
                                                int32_t    block_width,