@nativescript/visionos 8.6.0 → 8.7.0-rc.0

package/framework/internal/{metadata-generator-arm64/bin/lib/clang/13.0.1 → metadata-generator-x86_64/bin/lib/clang/15.0.7}/include/avx2intrin.h

@@ -20,25 +20,25 @@
 
 /* SSE4 Multiple Packed Sums of Absolute Difference.  */
 #define _mm256_mpsadbw_epu8(X, Y, M) \
-  (__m256i)__builtin_ia32_mpsadbw256((__v32qi)(__m256i)(X), \
-                                     (__v32qi)(__m256i)(Y), (int)(M))
+  ((__m256i)__builtin_ia32_mpsadbw256((__v32qi)(__m256i)(X), \
+                                      (__v32qi)(__m256i)(Y), (int)(M)))
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_abs_epi8(__m256i __a)
 {
-    return (__m256i)__builtin_ia32_pabsb256((__v32qi)__a);
+    return (__m256i)__builtin_elementwise_abs((__v32qs)__a);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_abs_epi16(__m256i __a)
 {
-    return (__m256i)__builtin_ia32_pabsw256((__v16hi)__a);
+    return (__m256i)__builtin_elementwise_abs((__v16hi)__a);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_abs_epi32(__m256i __a)
 {
-    return (__m256i)__builtin_ia32_pabsd256((__v8si)__a);
+    return (__m256i)__builtin_elementwise_abs((__v8si)__a);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
@@ -92,30 +92,30 @@ _mm256_add_epi64(__m256i __a, __m256i __b)
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_adds_epi8(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_paddsb256((__v32qi)__a, (__v32qi)__b);
+  return (__m256i)__builtin_elementwise_add_sat((__v32qs)__a, (__v32qs)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_adds_epi16(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_paddsw256((__v16hi)__a, (__v16hi)__b);
+  return (__m256i)__builtin_elementwise_add_sat((__v16hi)__a, (__v16hi)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_adds_epu8(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_paddusb256((__v32qi)__a, (__v32qi)__b);
+  return (__m256i)__builtin_elementwise_add_sat((__v32qu)__a, (__v32qu)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_adds_epu16(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_paddusw256((__v16hi)__a, (__v16hi)__b);
+  return (__m256i)__builtin_elementwise_add_sat((__v16hu)__a, (__v16hu)__b);
 }
 
 #define _mm256_alignr_epi8(a, b, n) \
-  (__m256i)__builtin_ia32_palignr256((__v32qi)(__m256i)(a), \
-                                     (__v32qi)(__m256i)(b), (n))
+  ((__m256i)__builtin_ia32_palignr256((__v32qi)(__m256i)(a), \
+                                      (__v32qi)(__m256i)(b), (n)))
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_and_si256(__m256i __a, __m256i __b)
@@ -149,8 +149,8 @@ _mm256_blendv_epi8(__m256i __V1, __m256i __V2, __m256i __M)
 }
 
 #define _mm256_blend_epi16(V1, V2, M) \
-  (__m256i)__builtin_ia32_pblendw256((__v16hi)(__m256i)(V1), \
-                                     (__v16hi)(__m256i)(V2), (int)(M))
+  ((__m256i)__builtin_ia32_pblendw256((__v16hi)(__m256i)(V1), \
+                                      (__v16hi)(__m256i)(V2), (int)(M)))
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_cmpeq_epi8(__m256i __a, __m256i __b)
@@ -253,73 +253,73 @@ _mm256_madd_epi16(__m256i __a, __m256i __b)
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_max_epi8(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_pmaxsb256((__v32qi)__a, (__v32qi)__b);
+  return (__m256i)__builtin_elementwise_max((__v32qs)__a, (__v32qs)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_max_epi16(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_pmaxsw256((__v16hi)__a, (__v16hi)__b);
+  return (__m256i)__builtin_elementwise_max((__v16hi)__a, (__v16hi)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_max_epi32(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_pmaxsd256((__v8si)__a, (__v8si)__b);
+  return (__m256i)__builtin_elementwise_max((__v8si)__a, (__v8si)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_max_epu8(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_pmaxub256((__v32qi)__a, (__v32qi)__b);
+  return (__m256i)__builtin_elementwise_max((__v32qu)__a, (__v32qu)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_max_epu16(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_pmaxuw256((__v16hi)__a, (__v16hi)__b);
+  return (__m256i)__builtin_elementwise_max((__v16hu)__a, (__v16hu)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_max_epu32(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_pmaxud256((__v8si)__a, (__v8si)__b);
+  return (__m256i)__builtin_elementwise_max((__v8su)__a, (__v8su)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_min_epi8(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_pminsb256((__v32qi)__a, (__v32qi)__b);
+  return (__m256i)__builtin_elementwise_min((__v32qs)__a, (__v32qs)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_min_epi16(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_pminsw256((__v16hi)__a, (__v16hi)__b);
+  return (__m256i)__builtin_elementwise_min((__v16hi)__a, (__v16hi)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_min_epi32(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_pminsd256((__v8si)__a, (__v8si)__b);
+  return (__m256i)__builtin_elementwise_min((__v8si)__a, (__v8si)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_min_epu8(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_pminub256((__v32qi)__a, (__v32qi)__b);
+  return (__m256i)__builtin_elementwise_min((__v32qu)__a, (__v32qu)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_min_epu16(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_pminuw256 ((__v16hi)__a, (__v16hi)__b);
+  return (__m256i)__builtin_elementwise_min((__v16hu)__a, (__v16hu)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_min_epu32(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_pminud256((__v8si)__a, (__v8si)__b);
+  return (__m256i)__builtin_elementwise_min((__v8su)__a, (__v8su)__b);
 }
 
 static __inline__ int __DEFAULT_FN_ATTRS256
@@ -467,13 +467,13 @@ _mm256_shuffle_epi8(__m256i __a, __m256i __b)
 }
 
 #define _mm256_shuffle_epi32(a, imm) \
-  (__m256i)__builtin_ia32_pshufd256((__v8si)(__m256i)(a), (int)(imm))
+  ((__m256i)__builtin_ia32_pshufd256((__v8si)(__m256i)(a), (int)(imm)))
 
 #define _mm256_shufflehi_epi16(a, imm) \
-  (__m256i)__builtin_ia32_pshufhw256((__v16hi)(__m256i)(a), (int)(imm))
+  ((__m256i)__builtin_ia32_pshufhw256((__v16hi)(__m256i)(a), (int)(imm)))
 
 #define _mm256_shufflelo_epi16(a, imm) \
-  (__m256i)__builtin_ia32_pshuflw256((__v16hi)(__m256i)(a), (int)(imm))
+  ((__m256i)__builtin_ia32_pshuflw256((__v16hi)(__m256i)(a), (int)(imm)))
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_sign_epi8(__m256i __a, __m256i __b)
@@ -494,10 +494,10 @@ _mm256_sign_epi32(__m256i __a, __m256i __b)
 }
 
 #define _mm256_slli_si256(a, imm) \
-  (__m256i)__builtin_ia32_pslldqi256_byteshift((__v4di)(__m256i)(a), (int)(imm))
+  ((__m256i)__builtin_ia32_pslldqi256_byteshift((__v4di)(__m256i)(a), (int)(imm)))
 
 #define _mm256_bslli_epi128(a, imm) \
-  (__m256i)__builtin_ia32_pslldqi256_byteshift((__v4di)(__m256i)(a), (int)(imm))
+  ((__m256i)__builtin_ia32_pslldqi256_byteshift((__v4di)(__m256i)(a), (int)(imm)))
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_slli_epi16(__m256i __a, int __count)
@@ -560,10 +560,10 @@ _mm256_sra_epi32(__m256i __a, __m128i __count)
 }
 
 #define _mm256_srli_si256(a, imm) \
-  (__m256i)__builtin_ia32_psrldqi256_byteshift((__m256i)(a), (int)(imm))
+  ((__m256i)__builtin_ia32_psrldqi256_byteshift((__m256i)(a), (int)(imm)))
 
 #define _mm256_bsrli_epi128(a, imm) \
-  (__m256i)__builtin_ia32_psrldqi256_byteshift((__m256i)(a), (int)(imm))
+  ((__m256i)__builtin_ia32_psrldqi256_byteshift((__m256i)(a), (int)(imm)))
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_srli_epi16(__m256i __a, int __count)
@@ -628,25 +628,25 @@ _mm256_sub_epi64(__m256i __a, __m256i __b)
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_subs_epi8(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_psubsb256((__v32qi)__a, (__v32qi)__b);
+  return (__m256i)__builtin_elementwise_sub_sat((__v32qs)__a, (__v32qs)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_subs_epi16(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_psubsw256((__v16hi)__a, (__v16hi)__b);
+  return (__m256i)__builtin_elementwise_sub_sat((__v16hi)__a, (__v16hi)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_subs_epu8(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_psubusb256((__v32qi)__a, (__v32qi)__b);
+  return (__m256i)__builtin_elementwise_sub_sat((__v32qu)__a, (__v32qu)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_subs_epu16(__m256i __a, __m256i __b)
 {
-  return (__m256i)__builtin_ia32_psubusw256((__v16hi)__a, (__v16hi)__b);
+  return (__m256i)__builtin_elementwise_sub_sat((__v16hu)__a, (__v16hu)__b);
 }
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
@@ -743,12 +743,12 @@ _mm256_broadcastsi128_si256(__m128i __X)
 #define _mm_broadcastsi128_si256(X) _mm256_broadcastsi128_si256(X)
 
 #define _mm_blend_epi32(V1, V2, M) \
-  (__m128i)__builtin_ia32_pblendd128((__v4si)(__m128i)(V1), \
-                                     (__v4si)(__m128i)(V2), (int)(M))
+  ((__m128i)__builtin_ia32_pblendd128((__v4si)(__m128i)(V1), \
+                                      (__v4si)(__m128i)(V2), (int)(M)))
 
 #define _mm256_blend_epi32(V1, V2, M) \
-  (__m256i)__builtin_ia32_pblendd256((__v8si)(__m256i)(V1), \
-                                     (__v8si)(__m256i)(V2), (int)(M))
+  ((__m256i)__builtin_ia32_pblendd256((__v8si)(__m256i)(V1), \
+                                      (__v8si)(__m256i)(V2), (int)(M)))
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_broadcastb_epi8(__m128i __X)
@@ -806,7 +806,7 @@ _mm256_permutevar8x32_epi32(__m256i __a, __m256i __b)
 }
 
 #define _mm256_permute4x64_pd(V, M) \
-  (__m256d)__builtin_ia32_permdf256((__v4df)(__m256d)(V), (int)(M))
+  ((__m256d)__builtin_ia32_permdf256((__v4df)(__m256d)(V), (int)(M)))
 
 static __inline__ __m256 __DEFAULT_FN_ATTRS256
 _mm256_permutevar8x32_ps(__m256 __a, __m256i __b)
@@ -815,17 +815,17 @@ _mm256_permutevar8x32_ps(__m256 __a, __m256i __b)
 }
 
 #define _mm256_permute4x64_epi64(V, M) \
-  (__m256i)__builtin_ia32_permdi256((__v4di)(__m256i)(V), (int)(M))
+  ((__m256i)__builtin_ia32_permdi256((__v4di)(__m256i)(V), (int)(M)))
 
 #define _mm256_permute2x128_si256(V1, V2, M) \
-  (__m256i)__builtin_ia32_permti256((__m256i)(V1), (__m256i)(V2), (int)(M))
+  ((__m256i)__builtin_ia32_permti256((__m256i)(V1), (__m256i)(V2), (int)(M)))
 
 #define _mm256_extracti128_si256(V, M) \
-  (__m128i)__builtin_ia32_extract128i256((__v4di)(__m256i)(V), (int)(M))
+  ((__m128i)__builtin_ia32_extract128i256((__v4di)(__m256i)(V), (int)(M)))
 
 #define _mm256_inserti128_si256(V1, V2, M) \
-  (__m256i)__builtin_ia32_insert128i256((__v4di)(__m256i)(V1), \
-                                        (__v2di)(__m128i)(V2), (int)(M))
+  ((__m256i)__builtin_ia32_insert128i256((__v4di)(__m256i)(V1), \
+                                         (__v2di)(__m128i)(V2), (int)(M)))
 
 static __inline__ __m256i __DEFAULT_FN_ATTRS256
 _mm256_maskload_epi32(int const *__X, __m256i __M)
@@ -936,211 +936,211 @@ _mm_srlv_epi64(__m128i __X, __m128i __Y)
 }
 
 #define _mm_mask_i32gather_pd(a, m, i, mask, s) \
-  (__m128d)__builtin_ia32_gatherd_pd((__v2df)(__m128i)(a), \
-                                     (double const *)(m), \
-                                     (__v4si)(__m128i)(i), \
-                                     (__v2df)(__m128d)(mask), (s))
+  ((__m128d)__builtin_ia32_gatherd_pd((__v2df)(__m128i)(a), \
+                                      (double const *)(m), \
+                                      (__v4si)(__m128i)(i), \
+                                      (__v2df)(__m128d)(mask), (s)))
 
 #define _mm256_mask_i32gather_pd(a, m, i, mask, s) \
-  (__m256d)__builtin_ia32_gatherd_pd256((__v4df)(__m256d)(a), \
-                                        (double const *)(m), \
-                                        (__v4si)(__m128i)(i), \
-                                        (__v4df)(__m256d)(mask), (s))
+  ((__m256d)__builtin_ia32_gatherd_pd256((__v4df)(__m256d)(a), \
+                                         (double const *)(m), \
+                                         (__v4si)(__m128i)(i), \
+                                         (__v4df)(__m256d)(mask), (s)))
 
 #define _mm_mask_i64gather_pd(a, m, i, mask, s) \
-  (__m128d)__builtin_ia32_gatherq_pd((__v2df)(__m128d)(a), \
-                                     (double const *)(m), \
-                                     (__v2di)(__m128i)(i), \
-                                     (__v2df)(__m128d)(mask), (s))
+  ((__m128d)__builtin_ia32_gatherq_pd((__v2df)(__m128d)(a), \
+                                      (double const *)(m), \
+                                      (__v2di)(__m128i)(i), \
+                                      (__v2df)(__m128d)(mask), (s)))
 
 #define _mm256_mask_i64gather_pd(a, m, i, mask, s) \
-  (__m256d)__builtin_ia32_gatherq_pd256((__v4df)(__m256d)(a), \
-                                        (double const *)(m), \
-                                        (__v4di)(__m256i)(i), \
-                                        (__v4df)(__m256d)(mask), (s))
+  ((__m256d)__builtin_ia32_gatherq_pd256((__v4df)(__m256d)(a), \
+                                         (double const *)(m), \
+                                         (__v4di)(__m256i)(i), \
+                                         (__v4df)(__m256d)(mask), (s)))
 
 #define _mm_mask_i32gather_ps(a, m, i, mask, s) \
-  (__m128)__builtin_ia32_gatherd_ps((__v4sf)(__m128)(a), \
-                                    (float const *)(m), \
-                                    (__v4si)(__m128i)(i), \
-                                    (__v4sf)(__m128)(mask), (s))
+  ((__m128)__builtin_ia32_gatherd_ps((__v4sf)(__m128)(a), \
+                                     (float const *)(m), \
+                                     (__v4si)(__m128i)(i), \
+                                     (__v4sf)(__m128)(mask), (s)))
 
 #define _mm256_mask_i32gather_ps(a, m, i, mask, s) \
-  (__m256)__builtin_ia32_gatherd_ps256((__v8sf)(__m256)(a), \
-                                       (float const *)(m), \
-                                       (__v8si)(__m256i)(i), \
-                                       (__v8sf)(__m256)(mask), (s))
+  ((__m256)__builtin_ia32_gatherd_ps256((__v8sf)(__m256)(a), \
+                                        (float const *)(m), \
+                                        (__v8si)(__m256i)(i), \
+                                        (__v8sf)(__m256)(mask), (s)))
 
 #define _mm_mask_i64gather_ps(a, m, i, mask, s) \
-  (__m128)__builtin_ia32_gatherq_ps((__v4sf)(__m128)(a), \
-                                    (float const *)(m), \
-                                    (__v2di)(__m128i)(i), \
-                                    (__v4sf)(__m128)(mask), (s))
+  ((__m128)__builtin_ia32_gatherq_ps((__v4sf)(__m128)(a), \
+                                     (float const *)(m), \
+                                     (__v2di)(__m128i)(i), \
+                                     (__v4sf)(__m128)(mask), (s)))
 
 #define _mm256_mask_i64gather_ps(a, m, i, mask, s) \
-  (__m128)__builtin_ia32_gatherq_ps256((__v4sf)(__m128)(a), \
-                                       (float const *)(m), \
-                                       (__v4di)(__m256i)(i), \
-                                       (__v4sf)(__m128)(mask), (s))
+  ((__m128)__builtin_ia32_gatherq_ps256((__v4sf)(__m128)(a), \
+                                        (float const *)(m), \
+                                        (__v4di)(__m256i)(i), \
+                                        (__v4sf)(__m128)(mask), (s)))
 
 #define _mm_mask_i32gather_epi32(a, m, i, mask, s) \
-  (__m128i)__builtin_ia32_gatherd_d((__v4si)(__m128i)(a), \
-                                    (int const *)(m), \
-                                    (__v4si)(__m128i)(i), \
-                                    (__v4si)(__m128i)(mask), (s))
+  ((__m128i)__builtin_ia32_gatherd_d((__v4si)(__m128i)(a), \
+                                     (int const *)(m), \
+                                     (__v4si)(__m128i)(i), \
+                                     (__v4si)(__m128i)(mask), (s)))
 
 #define _mm256_mask_i32gather_epi32(a, m, i, mask, s) \
-  (__m256i)__builtin_ia32_gatherd_d256((__v8si)(__m256i)(a), \
-                                       (int const *)(m), \
-                                       (__v8si)(__m256i)(i), \
-                                       (__v8si)(__m256i)(mask), (s))
+  ((__m256i)__builtin_ia32_gatherd_d256((__v8si)(__m256i)(a), \
+                                        (int const *)(m), \
+                                        (__v8si)(__m256i)(i), \
+                                        (__v8si)(__m256i)(mask), (s)))
 
 #define _mm_mask_i64gather_epi32(a, m, i, mask, s) \
-  (__m128i)__builtin_ia32_gatherq_d((__v4si)(__m128i)(a), \
-                                    (int const *)(m), \
-                                    (__v2di)(__m128i)(i), \
-                                    (__v4si)(__m128i)(mask), (s))
+  ((__m128i)__builtin_ia32_gatherq_d((__v4si)(__m128i)(a), \
+                                     (int const *)(m), \
+                                     (__v2di)(__m128i)(i), \
+                                     (__v4si)(__m128i)(mask), (s)))
 
 #define _mm256_mask_i64gather_epi32(a, m, i, mask, s) \
-  (__m128i)__builtin_ia32_gatherq_d256((__v4si)(__m128i)(a), \
-                                       (int const *)(m), \
-                                       (__v4di)(__m256i)(i), \
-                                       (__v4si)(__m128i)(mask), (s))
+  ((__m128i)__builtin_ia32_gatherq_d256((__v4si)(__m128i)(a), \
+                                        (int const *)(m), \
+                                        (__v4di)(__m256i)(i), \
+                                        (__v4si)(__m128i)(mask), (s)))
 
 #define _mm_mask_i32gather_epi64(a, m, i, mask, s) \
-  (__m128i)__builtin_ia32_gatherd_q((__v2di)(__m128i)(a), \
-                                    (long long const *)(m), \
-                                    (__v4si)(__m128i)(i), \
-                                    (__v2di)(__m128i)(mask), (s))
+  ((__m128i)__builtin_ia32_gatherd_q((__v2di)(__m128i)(a), \
+                                     (long long const *)(m), \
+                                     (__v4si)(__m128i)(i), \
+                                     (__v2di)(__m128i)(mask), (s)))
 
 #define _mm256_mask_i32gather_epi64(a, m, i, mask, s) \
-  (__m256i)__builtin_ia32_gatherd_q256((__v4di)(__m256i)(a), \
-                                       (long long const *)(m), \
-                                       (__v4si)(__m128i)(i), \
-                                       (__v4di)(__m256i)(mask), (s))
+  ((__m256i)__builtin_ia32_gatherd_q256((__v4di)(__m256i)(a), \
+                                        (long long const *)(m), \
+                                        (__v4si)(__m128i)(i), \
+                                        (__v4di)(__m256i)(mask), (s)))
 
 #define _mm_mask_i64gather_epi64(a, m, i, mask, s) \
-  (__m128i)__builtin_ia32_gatherq_q((__v2di)(__m128i)(a), \
-                                    (long long const *)(m), \
-                                    (__v2di)(__m128i)(i), \
-                                    (__v2di)(__m128i)(mask), (s))
+  ((__m128i)__builtin_ia32_gatherq_q((__v2di)(__m128i)(a), \
+                                     (long long const *)(m), \
+                                     (__v2di)(__m128i)(i), \
+                                     (__v2di)(__m128i)(mask), (s)))
 
 #define _mm256_mask_i64gather_epi64(a, m, i, mask, s) \
-  (__m256i)__builtin_ia32_gatherq_q256((__v4di)(__m256i)(a), \
-                                       (long long const *)(m), \
-                                       (__v4di)(__m256i)(i), \
-                                       (__v4di)(__m256i)(mask), (s))
+  ((__m256i)__builtin_ia32_gatherq_q256((__v4di)(__m256i)(a), \
+                                        (long long const *)(m), \
+                                        (__v4di)(__m256i)(i), \
+                                        (__v4di)(__m256i)(mask), (s)))
 
 #define _mm_i32gather_pd(m, i, s) \
-  (__m128d)__builtin_ia32_gatherd_pd((__v2df)_mm_undefined_pd(), \
-                                     (double const *)(m), \
-                                     (__v4si)(__m128i)(i), \
-                                     (__v2df)_mm_cmpeq_pd(_mm_setzero_pd(), \
-                                                          _mm_setzero_pd()), \
-                                     (s))
+  ((__m128d)__builtin_ia32_gatherd_pd((__v2df)_mm_undefined_pd(), \
+                                      (double const *)(m), \
+                                      (__v4si)(__m128i)(i), \
+                                      (__v2df)_mm_cmpeq_pd(_mm_setzero_pd(), \
+                                                           _mm_setzero_pd()), \
+                                      (s)))
 
 #define _mm256_i32gather_pd(m, i, s) \
-  (__m256d)__builtin_ia32_gatherd_pd256((__v4df)_mm256_undefined_pd(), \
-                                        (double const *)(m), \
-                                        (__v4si)(__m128i)(i), \
-                                        (__v4df)_mm256_cmp_pd(_mm256_setzero_pd(), \
-                                                              _mm256_setzero_pd(), \
-                                                              _CMP_EQ_OQ), \
-                                        (s))
+  ((__m256d)__builtin_ia32_gatherd_pd256((__v4df)_mm256_undefined_pd(), \
+                                         (double const *)(m), \
+                                         (__v4si)(__m128i)(i), \
+                                         (__v4df)_mm256_cmp_pd(_mm256_setzero_pd(), \
+                                                               _mm256_setzero_pd(), \
+                                                               _CMP_EQ_OQ), \
+                                         (s)))
 
 #define _mm_i64gather_pd(m, i, s) \
-  (__m128d)__builtin_ia32_gatherq_pd((__v2df)_mm_undefined_pd(), \
-                                     (double const *)(m), \
-                                     (__v2di)(__m128i)(i), \
-                                     (__v2df)_mm_cmpeq_pd(_mm_setzero_pd(), \
-                                                          _mm_setzero_pd()), \
-                                     (s))
+  ((__m128d)__builtin_ia32_gatherq_pd((__v2df)_mm_undefined_pd(), \
+                                      (double const *)(m), \
+                                      (__v2di)(__m128i)(i), \
+                                      (__v2df)_mm_cmpeq_pd(_mm_setzero_pd(), \
+                                                           _mm_setzero_pd()), \
+                                      (s)))
 
 #define _mm256_i64gather_pd(m, i, s) \
-  (__m256d)__builtin_ia32_gatherq_pd256((__v4df)_mm256_undefined_pd(), \
-                                        (double const *)(m), \
-                                        (__v4di)(__m256i)(i), \
-                                        (__v4df)_mm256_cmp_pd(_mm256_setzero_pd(), \
-                                                              _mm256_setzero_pd(), \
-                                                              _CMP_EQ_OQ), \
-                                        (s))
+  ((__m256d)__builtin_ia32_gatherq_pd256((__v4df)_mm256_undefined_pd(), \
+                                         (double const *)(m), \
+                                         (__v4di)(__m256i)(i), \
+                                         (__v4df)_mm256_cmp_pd(_mm256_setzero_pd(), \
+                                                               _mm256_setzero_pd(), \
+                                                               _CMP_EQ_OQ), \
+                                         (s)))
 
 #define _mm_i32gather_ps(m, i, s) \
-  (__m128)__builtin_ia32_gatherd_ps((__v4sf)_mm_undefined_ps(), \
-                                    (float const *)(m), \
-                                    (__v4si)(__m128i)(i), \
-                                    (__v4sf)_mm_cmpeq_ps(_mm_setzero_ps(), \
-                                                         _mm_setzero_ps()), \
-                                    (s))
+  ((__m128)__builtin_ia32_gatherd_ps((__v4sf)_mm_undefined_ps(), \
+                                     (float const *)(m), \
+                                     (__v4si)(__m128i)(i), \
+                                     (__v4sf)_mm_cmpeq_ps(_mm_setzero_ps(), \
+                                                          _mm_setzero_ps()), \
+                                     (s)))
 
 #define _mm256_i32gather_ps(m, i, s) \
-  (__m256)__builtin_ia32_gatherd_ps256((__v8sf)_mm256_undefined_ps(), \
-                                       (float const *)(m), \
-                                       (__v8si)(__m256i)(i), \
-                                       (__v8sf)_mm256_cmp_ps(_mm256_setzero_ps(), \
-                                                             _mm256_setzero_ps(), \
-                                                             _CMP_EQ_OQ), \
-                                       (s))
+  ((__m256)__builtin_ia32_gatherd_ps256((__v8sf)_mm256_undefined_ps(), \
+                                        (float const *)(m), \
+                                        (__v8si)(__m256i)(i), \
+                                        (__v8sf)_mm256_cmp_ps(_mm256_setzero_ps(), \
+                                                              _mm256_setzero_ps(), \
+                                                              _CMP_EQ_OQ), \
+                                        (s)))
 
 #define _mm_i64gather_ps(m, i, s) \
-  (__m128)__builtin_ia32_gatherq_ps((__v4sf)_mm_undefined_ps(), \
-                                    (float const *)(m), \
-                                    (__v2di)(__m128i)(i), \
-                                    (__v4sf)_mm_cmpeq_ps(_mm_setzero_ps(), \
-                                                         _mm_setzero_ps()), \
-                                    (s))
+  ((__m128)__builtin_ia32_gatherq_ps((__v4sf)_mm_undefined_ps(), \
+                                     (float const *)(m), \
+                                     (__v2di)(__m128i)(i), \
+                                     (__v4sf)_mm_cmpeq_ps(_mm_setzero_ps(), \
+                                                          _mm_setzero_ps()), \
+                                     (s)))
 
 #define _mm256_i64gather_ps(m, i, s) \
-  (__m128)__builtin_ia32_gatherq_ps256((__v4sf)_mm_undefined_ps(), \
-                                       (float const *)(m), \
-                                       (__v4di)(__m256i)(i), \
-                                       (__v4sf)_mm_cmpeq_ps(_mm_setzero_ps(), \
-                                                            _mm_setzero_ps()), \
-                                       (s))
+  ((__m128)__builtin_ia32_gatherq_ps256((__v4sf)_mm_undefined_ps(), \
+                                        (float const *)(m), \
+                                        (__v4di)(__m256i)(i), \
+                                        (__v4sf)_mm_cmpeq_ps(_mm_setzero_ps(), \
+                                                             _mm_setzero_ps()), \
+                                        (s)))
 
 #define _mm_i32gather_epi32(m, i, s) \
-  (__m128i)__builtin_ia32_gatherd_d((__v4si)_mm_undefined_si128(), \
-                                    (int const *)(m), (__v4si)(__m128i)(i), \
-                                    (__v4si)_mm_set1_epi32(-1), (s))
+  ((__m128i)__builtin_ia32_gatherd_d((__v4si)_mm_undefined_si128(), \
+                                     (int const *)(m), (__v4si)(__m128i)(i), \
+                                     (__v4si)_mm_set1_epi32(-1), (s)))
 
 #define _mm256_i32gather_epi32(m, i, s) \
-  (__m256i)__builtin_ia32_gatherd_d256((__v8si)_mm256_undefined_si256(), \
-                                       (int const *)(m), (__v8si)(__m256i)(i), \
-                                       (__v8si)_mm256_set1_epi32(-1), (s))
+  ((__m256i)__builtin_ia32_gatherd_d256((__v8si)_mm256_undefined_si256(), \
+                                        (int const *)(m), (__v8si)(__m256i)(i), \
+                                        (__v8si)_mm256_set1_epi32(-1), (s)))
 
 #define _mm_i64gather_epi32(m, i, s) \
-  (__m128i)__builtin_ia32_gatherq_d((__v4si)_mm_undefined_si128(), \
-                                    (int const *)(m), (__v2di)(__m128i)(i), \
-                                    (__v4si)_mm_set1_epi32(-1), (s))
+  ((__m128i)__builtin_ia32_gatherq_d((__v4si)_mm_undefined_si128(), \
+                                     (int const *)(m), (__v2di)(__m128i)(i), \
+                                     (__v4si)_mm_set1_epi32(-1), (s)))
 
 #define _mm256_i64gather_epi32(m, i, s) \
-  (__m128i)__builtin_ia32_gatherq_d256((__v4si)_mm_undefined_si128(), \
-                                       (int const *)(m), (__v4di)(__m256i)(i), \
-                                       (__v4si)_mm_set1_epi32(-1), (s))
+  ((__m128i)__builtin_ia32_gatherq_d256((__v4si)_mm_undefined_si128(), \
+                                        (int const *)(m), (__v4di)(__m256i)(i), \
+                                        (__v4si)_mm_set1_epi32(-1), (s)))
 
 #define _mm_i32gather_epi64(m, i, s) \
-  (__m128i)__builtin_ia32_gatherd_q((__v2di)_mm_undefined_si128(), \
-                                    (long long const *)(m), \
-                                    (__v4si)(__m128i)(i), \
-                                    (__v2di)_mm_set1_epi64x(-1), (s))
+  ((__m128i)__builtin_ia32_gatherd_q((__v2di)_mm_undefined_si128(), \
+                                     (long long const *)(m), \
+                                     (__v4si)(__m128i)(i), \
+                                     (__v2di)_mm_set1_epi64x(-1), (s)))
 
 #define _mm256_i32gather_epi64(m, i, s) \
-  (__m256i)__builtin_ia32_gatherd_q256((__v4di)_mm256_undefined_si256(), \
-                                       (long long const *)(m), \
-                                       (__v4si)(__m128i)(i), \
-                                       (__v4di)_mm256_set1_epi64x(-1), (s))
+  ((__m256i)__builtin_ia32_gatherd_q256((__v4di)_mm256_undefined_si256(), \
+                                        (long long const *)(m), \
+                                        (__v4si)(__m128i)(i), \
+                                        (__v4di)_mm256_set1_epi64x(-1), (s)))
 
 #define _mm_i64gather_epi64(m, i, s) \
-  (__m128i)__builtin_ia32_gatherq_q((__v2di)_mm_undefined_si128(), \
-                                    (long long const *)(m), \
-                                    (__v2di)(__m128i)(i), \
-                                    (__v2di)_mm_set1_epi64x(-1), (s))
+  ((__m128i)__builtin_ia32_gatherq_q((__v2di)_mm_undefined_si128(), \
+                                     (long long const *)(m), \
+                                     (__v2di)(__m128i)(i), \
+                                     (__v2di)_mm_set1_epi64x(-1), (s)))
 
 #define _mm256_i64gather_epi64(m, i, s) \
-  (__m256i)__builtin_ia32_gatherq_q256((__v4di)_mm256_undefined_si256(), \
-                                       (long long const *)(m), \
-                                       (__v4di)(__m256i)(i), \
-                                       (__v4di)_mm256_set1_epi64x(-1), (s))
+  ((__m256i)__builtin_ia32_gatherq_q256((__v4di)_mm256_undefined_si256(), \
+                                        (long long const *)(m), \
+                                        (__v4di)(__m256i)(i), \
+                                        (__v4di)_mm256_set1_epi64x(-1), (s)))
 
 #undef __DEFAULT_FN_ATTRS256
 #undef __DEFAULT_FN_ATTRS128