@cpp-linter/clang-tidy-win32-x64 13.0.0 → 14.0.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (175) hide show
  1. package/README.md +2 -2
  2. package/bin/clang-tidy.exe +2 -2
  3. package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_cuda_complex_builtins.h +3 -3
  4. package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_cuda_intrinsics.h +32 -0
  5. package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_cuda_libdevice_declares.h +6 -0
  6. package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_cuda_math.h +1 -1
  7. package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_cuda_runtime_wrapper.h +62 -7
  8. package/lib/clang/14.0.0/include/__clang_cuda_texture_intrinsics.h +740 -0
  9. package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_hip_runtime_wrapper.h +17 -4
  10. package/lib/clang/{13.0.0 → 14.0.0}/include/__wmmintrin_aes.h +1 -1
  11. package/lib/clang/{13.0.0 → 14.0.0}/include/altivec.h +202 -92
  12. package/lib/clang/{13.0.0 → 14.0.0}/include/ammintrin.h +4 -0
  13. package/lib/clang/{13.0.0 → 14.0.0}/include/amxintrin.h +17 -16
  14. package/lib/clang/{13.0.0 → 14.0.0}/include/arm_acle.h +6 -0
  15. package/lib/clang/{13.0.0 → 14.0.0}/include/arm_neon.h +4467 -4469
  16. package/lib/clang/14.0.0/include/arm_neon_sve_bridge.h +184 -0
  17. package/lib/clang/{13.0.0 → 14.0.0}/include/avx2intrin.h +181 -181
  18. package/lib/clang/{13.0.0 → 14.0.0}/include/avx512bf16intrin.h +3 -3
  19. package/lib/clang/{13.0.0 → 14.0.0}/include/avx512bwintrin.h +71 -71
  20. package/lib/clang/{13.0.0 → 14.0.0}/include/avx512dqintrin.h +365 -365
  21. package/lib/clang/{13.0.0 → 14.0.0}/include/avx512erintrin.h +102 -102
  22. package/lib/clang/{13.0.0 → 14.0.0}/include/avx512fintrin.h +1573 -1569
  23. package/lib/clang/14.0.0/include/avx512fp16intrin.h +3349 -0
  24. package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vbmi2intrin.h +48 -48
  25. package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vlbf16intrin.h +51 -3
  26. package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vlbwintrin.h +94 -94
  27. package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vldqintrin.h +134 -134
  28. package/lib/clang/14.0.0/include/avx512vlfp16intrin.h +2068 -0
  29. package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vlintrin.h +563 -563
  30. package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vlvbmi2intrin.h +96 -96
  31. package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vlvnniintrin.h +8 -8
  32. package/lib/clang/{13.0.0 → 14.0.0}/include/avxintrin.h +189 -192
  33. package/lib/clang/{13.0.0 → 14.0.0}/include/cetintrin.h +10 -0
  34. package/lib/clang/{13.0.0 → 14.0.0}/include/cpuid.h +2 -1
  35. package/lib/clang/14.0.0/include/crc32intrin.h +100 -0
  36. package/lib/clang/{13.0.0 → 14.0.0}/include/emmintrin.h +21 -17
  37. package/lib/clang/{13.0.0 → 14.0.0}/include/f16cintrin.h +4 -4
  38. package/lib/clang/{13.0.0 → 14.0.0}/include/float.h +15 -6
  39. package/lib/clang/{13.0.0 → 14.0.0}/include/gfniintrin.h +47 -48
  40. package/lib/clang/{13.0.0 → 14.0.0}/include/hexagon_protos.h +0 -11
  41. package/lib/clang/{13.0.0 → 14.0.0}/include/hexagon_types.h +0 -32
  42. package/lib/clang/{13.0.0 → 14.0.0}/include/hvx_hexagon_protos.h +1200 -409
  43. package/lib/clang/{13.0.0 → 14.0.0}/include/ia32intrin.h +6 -6
  44. package/lib/clang/{13.0.0 → 14.0.0}/include/immintrin.h +26 -8
  45. package/lib/clang/{13.0.0 → 14.0.0}/include/intrin.h +24 -16
  46. package/lib/clang/{13.0.0 → 14.0.0}/include/keylockerintrin.h +2 -8
  47. package/lib/clang/{13.0.0 → 14.0.0}/include/limits.h +20 -0
  48. package/lib/clang/{13.0.0 → 14.0.0}/include/mmintrin.h +4 -0
  49. package/lib/clang/{13.0.0 → 14.0.0}/include/nmmintrin.h +4 -0
  50. package/lib/clang/{13.0.0 → 14.0.0}/include/opencl-c-base.h +69 -11
  51. package/lib/clang/{13.0.0 → 14.0.0}/include/opencl-c.h +1699 -459
  52. package/lib/clang/{13.0.0 → 14.0.0}/include/openmp_wrappers/complex +10 -1
  53. package/lib/clang/{13.0.0 → 14.0.0}/include/openmp_wrappers/complex.h +9 -0
  54. package/lib/clang/{13.0.0 → 14.0.0}/include/pmmintrin.h +4 -0
  55. package/lib/clang/{13.0.0 → 14.0.0}/include/ppc_wrappers/emmintrin.h +3 -2
  56. package/lib/clang/{13.0.0 → 14.0.0}/include/ppc_wrappers/mm_malloc.h +2 -2
  57. package/lib/clang/{13.0.0 → 14.0.0}/include/ppc_wrappers/mmintrin.h +3 -2
  58. package/lib/clang/{13.0.0 → 14.0.0}/include/ppc_wrappers/pmmintrin.h +3 -2
  59. package/lib/clang/{13.0.0 → 14.0.0}/include/ppc_wrappers/smmintrin.h +4 -3
  60. package/lib/clang/{13.0.0 → 14.0.0}/include/ppc_wrappers/tmmintrin.h +3 -2
  61. package/lib/clang/{13.0.0 → 14.0.0}/include/ppc_wrappers/xmmintrin.h +3 -2
  62. package/lib/clang/{13.0.0 → 14.0.0}/include/prfchwintrin.h +5 -2
  63. package/lib/clang/14.0.0/include/riscv_vector.h +95423 -0
  64. package/lib/clang/{13.0.0 → 14.0.0}/include/smmintrin.h +80 -159
  65. package/lib/clang/{13.0.0 → 14.0.0}/include/stdatomic.h +14 -1
  66. package/lib/clang/{13.0.0 → 14.0.0}/include/stdint.h +168 -0
  67. package/lib/clang/{13.0.0 → 14.0.0}/include/tmmintrin.h +10 -6
  68. package/lib/clang/{13.0.0 → 14.0.0}/include/unwind.h +2 -1
  69. package/lib/clang/{13.0.0 → 14.0.0}/include/vaesintrin.h +1 -1
  70. package/lib/clang/{13.0.0 → 14.0.0}/include/vpclmulqdqintrin.h +6 -6
  71. package/lib/clang/{13.0.0 → 14.0.0}/include/wasm_simd128.h +167 -24
  72. package/lib/clang/{13.0.0 → 14.0.0}/include/wmmintrin.h +4 -0
  73. package/lib/clang/{13.0.0 → 14.0.0}/include/x86gprintrin.h +12 -0
  74. package/lib/clang/{13.0.0 → 14.0.0}/include/xmmintrin.h +9 -5
  75. package/lib/clang/{13.0.0 → 14.0.0}/include/xopintrin.h +31 -31
  76. package/package.json +6 -6
  77. package/lib/clang/13.0.0/include/riscv_vector.h +0 -123865
  78. /package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_cuda_builtin_vars.h +0 -0
  79. /package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_cuda_cmath.h +0 -0
  80. /package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_cuda_device_functions.h +0 -0
  81. /package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_cuda_math_forward_declares.h +0 -0
  82. /package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_hip_cmath.h +0 -0
  83. /package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_hip_libdevice_declares.h +0 -0
  84. /package/lib/clang/{13.0.0 → 14.0.0}/include/__clang_hip_math.h +0 -0
  85. /package/lib/clang/{13.0.0 → 14.0.0}/include/__stddef_max_align_t.h +0 -0
  86. /package/lib/clang/{13.0.0 → 14.0.0}/include/__wmmintrin_pclmul.h +0 -0
  87. /package/lib/clang/{13.0.0 → 14.0.0}/include/adxintrin.h +0 -0
  88. /package/lib/clang/{13.0.0 → 14.0.0}/include/arm64intr.h +0 -0
  89. /package/lib/clang/{13.0.0 → 14.0.0}/include/arm_bf16.h +0 -0
  90. /package/lib/clang/{13.0.0 → 14.0.0}/include/arm_cde.h +0 -0
  91. /package/lib/clang/{13.0.0 → 14.0.0}/include/arm_cmse.h +0 -0
  92. /package/lib/clang/{13.0.0 → 14.0.0}/include/arm_fp16.h +0 -0
  93. /package/lib/clang/{13.0.0 → 14.0.0}/include/arm_mve.h +0 -0
  94. /package/lib/clang/{13.0.0 → 14.0.0}/include/arm_sve.h +0 -0
  95. /package/lib/clang/{13.0.0 → 14.0.0}/include/armintr.h +0 -0
  96. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512bitalgintrin.h +0 -0
  97. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512cdintrin.h +0 -0
  98. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512ifmaintrin.h +0 -0
  99. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512ifmavlintrin.h +0 -0
  100. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512pfintrin.h +0 -0
  101. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vbmiintrin.h +0 -0
  102. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vbmivlintrin.h +0 -0
  103. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vlbitalgintrin.h +0 -0
  104. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vlcdintrin.h +0 -0
  105. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vlvp2intersectintrin.h +0 -0
  106. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vnniintrin.h +0 -0
  107. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vp2intersectintrin.h +0 -0
  108. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vpopcntdqintrin.h +0 -0
  109. /package/lib/clang/{13.0.0 → 14.0.0}/include/avx512vpopcntdqvlintrin.h +0 -0
  110. /package/lib/clang/{13.0.0 → 14.0.0}/include/avxvnniintrin.h +0 -0
  111. /package/lib/clang/{13.0.0 → 14.0.0}/include/bmi2intrin.h +0 -0
  112. /package/lib/clang/{13.0.0 → 14.0.0}/include/bmiintrin.h +0 -0
  113. /package/lib/clang/{13.0.0 → 14.0.0}/include/builtins.h +0 -0
  114. /package/lib/clang/{13.0.0 → 14.0.0}/include/cet.h +0 -0
  115. /package/lib/clang/{13.0.0 → 14.0.0}/include/cldemoteintrin.h +0 -0
  116. /package/lib/clang/{13.0.0 → 14.0.0}/include/clflushoptintrin.h +0 -0
  117. /package/lib/clang/{13.0.0 → 14.0.0}/include/clwbintrin.h +0 -0
  118. /package/lib/clang/{13.0.0 → 14.0.0}/include/clzerointrin.h +0 -0
  119. /package/lib/clang/{13.0.0 → 14.0.0}/include/cuda_wrappers/algorithm +0 -0
  120. /package/lib/clang/{13.0.0 → 14.0.0}/include/cuda_wrappers/complex +0 -0
  121. /package/lib/clang/{13.0.0 → 14.0.0}/include/cuda_wrappers/new +0 -0
  122. /package/lib/clang/{13.0.0 → 14.0.0}/include/enqcmdintrin.h +0 -0
  123. /package/lib/clang/{13.0.0 → 14.0.0}/include/fma4intrin.h +0 -0
  124. /package/lib/clang/{13.0.0 → 14.0.0}/include/fmaintrin.h +0 -0
  125. /package/lib/clang/{13.0.0 → 14.0.0}/include/fxsrintrin.h +0 -0
  126. /package/lib/clang/{13.0.0 → 14.0.0}/include/hexagon_circ_brev_intrinsics.h +0 -0
  127. /package/lib/clang/{13.0.0 → 14.0.0}/include/hresetintrin.h +0 -0
  128. /package/lib/clang/{13.0.0 → 14.0.0}/include/htmintrin.h +0 -0
  129. /package/lib/clang/{13.0.0 → 14.0.0}/include/htmxlintrin.h +0 -0
  130. /package/lib/clang/{13.0.0 → 14.0.0}/include/inttypes.h +0 -0
  131. /package/lib/clang/{13.0.0 → 14.0.0}/include/invpcidintrin.h +0 -0
  132. /package/lib/clang/{13.0.0 → 14.0.0}/include/iso646.h +0 -0
  133. /package/lib/clang/{13.0.0 → 14.0.0}/include/lwpintrin.h +0 -0
  134. /package/lib/clang/{13.0.0 → 14.0.0}/include/lzcntintrin.h +0 -0
  135. /package/lib/clang/{13.0.0 → 14.0.0}/include/mm3dnow.h +0 -0
  136. /package/lib/clang/{13.0.0 → 14.0.0}/include/mm_malloc.h +0 -0
  137. /package/lib/clang/{13.0.0 → 14.0.0}/include/module.modulemap +0 -0
  138. /package/lib/clang/{13.0.0 → 14.0.0}/include/movdirintrin.h +0 -0
  139. /package/lib/clang/{13.0.0 → 14.0.0}/include/msa.h +0 -0
  140. /package/lib/clang/{13.0.0 → 14.0.0}/include/mwaitxintrin.h +0 -0
  141. /package/lib/clang/{13.0.0 → 14.0.0}/include/openmp_wrappers/__clang_openmp_device_functions.h +0 -0
  142. /package/lib/clang/{13.0.0 → 14.0.0}/include/openmp_wrappers/cmath +0 -0
  143. /package/lib/clang/{13.0.0 → 14.0.0}/include/openmp_wrappers/complex_cmath.h +0 -0
  144. /package/lib/clang/{13.0.0 → 14.0.0}/include/openmp_wrappers/math.h +0 -0
  145. /package/lib/clang/{13.0.0 → 14.0.0}/include/openmp_wrappers/new +0 -0
  146. /package/lib/clang/{13.0.0 → 14.0.0}/include/pconfigintrin.h +0 -0
  147. /package/lib/clang/{13.0.0 → 14.0.0}/include/pkuintrin.h +0 -0
  148. /package/lib/clang/{13.0.0 → 14.0.0}/include/popcntintrin.h +0 -0
  149. /package/lib/clang/{13.0.0 → 14.0.0}/include/ptwriteintrin.h +0 -0
  150. /package/lib/clang/{13.0.0 → 14.0.0}/include/rdseedintrin.h +0 -0
  151. /package/lib/clang/{13.0.0 → 14.0.0}/include/rtmintrin.h +0 -0
  152. /package/lib/clang/{13.0.0 → 14.0.0}/include/s390intrin.h +0 -0
  153. /package/lib/clang/{13.0.0 → 14.0.0}/include/serializeintrin.h +0 -0
  154. /package/lib/clang/{13.0.0 → 14.0.0}/include/sgxintrin.h +0 -0
  155. /package/lib/clang/{13.0.0 → 14.0.0}/include/shaintrin.h +0 -0
  156. /package/lib/clang/{13.0.0 → 14.0.0}/include/stdalign.h +0 -0
  157. /package/lib/clang/{13.0.0 → 14.0.0}/include/stdarg.h +0 -0
  158. /package/lib/clang/{13.0.0 → 14.0.0}/include/stdbool.h +0 -0
  159. /package/lib/clang/{13.0.0 → 14.0.0}/include/stddef.h +0 -0
  160. /package/lib/clang/{13.0.0 → 14.0.0}/include/stdnoreturn.h +0 -0
  161. /package/lib/clang/{13.0.0 → 14.0.0}/include/tbmintrin.h +0 -0
  162. /package/lib/clang/{13.0.0 → 14.0.0}/include/tgmath.h +0 -0
  163. /package/lib/clang/{13.0.0 → 14.0.0}/include/tsxldtrkintrin.h +0 -0
  164. /package/lib/clang/{13.0.0 → 14.0.0}/include/uintrintrin.h +0 -0
  165. /package/lib/clang/{13.0.0 → 14.0.0}/include/vadefs.h +0 -0
  166. /package/lib/clang/{13.0.0 → 14.0.0}/include/varargs.h +0 -0
  167. /package/lib/clang/{13.0.0 → 14.0.0}/include/vecintrin.h +0 -0
  168. /package/lib/clang/{13.0.0 → 14.0.0}/include/waitpkgintrin.h +0 -0
  169. /package/lib/clang/{13.0.0 → 14.0.0}/include/wbnoinvdintrin.h +0 -0
  170. /package/lib/clang/{13.0.0 → 14.0.0}/include/x86intrin.h +0 -0
  171. /package/lib/clang/{13.0.0 → 14.0.0}/include/xsavecintrin.h +0 -0
  172. /package/lib/clang/{13.0.0 → 14.0.0}/include/xsaveintrin.h +0 -0
  173. /package/lib/clang/{13.0.0 → 14.0.0}/include/xsaveoptintrin.h +0 -0
  174. /package/lib/clang/{13.0.0 → 14.0.0}/include/xsavesintrin.h +0 -0
  175. /package/lib/clang/{13.0.0 → 14.0.0}/include/xtestintrin.h +0 -0
package/lib/clang/{13.0.0 → 14.0.0}/include/avx512fintrin.h RENAMED
@@ -26,6 +26,10 @@ typedef unsigned short __v32hu __attribute__((__vector_size__(64)));
26
26
  typedef unsigned long long __v8du __attribute__((__vector_size__(64)));
27
27
  typedef unsigned int __v16su __attribute__((__vector_size__(64)));
28
28
 
29
+ /* We need an explicitly signed variant for char. Note that this shouldn't
30
+ * appear in the interface though. */
31
+ typedef signed char __v64qs __attribute__((__vector_size__(64)));
32
+
29
33
  typedef float __m512 __attribute__((__vector_size__(64), __aligned__(64)));
30
34
  typedef double __m512d __attribute__((__vector_size__(64), __aligned__(64)));
31
35
  typedef long long __m512i __attribute__((__vector_size__(64), __aligned__(64)));
@@ -937,18 +941,18 @@ _mm512_maskz_sub_epi32(__mmask16 __U, __m512i __A, __m512i __B)
937
941
  }
938
942
 
939
943
  #define _mm512_max_round_pd(A, B, R) \
940
- (__m512d)__builtin_ia32_maxpd512((__v8df)(__m512d)(A), \
941
- (__v8df)(__m512d)(B), (int)(R))
944
+ ((__m512d)__builtin_ia32_maxpd512((__v8df)(__m512d)(A), \
945
+ (__v8df)(__m512d)(B), (int)(R)))
942
946
 
943
947
  #define _mm512_mask_max_round_pd(W, U, A, B, R) \
944
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
948
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
945
949
  (__v8df)_mm512_max_round_pd((A), (B), (R)), \
946
- (__v8df)(W))
950
+ (__v8df)(W)))
947
951
 
948
952
  #define _mm512_maskz_max_round_pd(U, A, B, R) \
949
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
953
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
950
954
  (__v8df)_mm512_max_round_pd((A), (B), (R)), \
951
- (__v8df)_mm512_setzero_pd())
955
+ (__v8df)_mm512_setzero_pd()))
952
956
 
953
957
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
954
958
  _mm512_max_pd(__m512d __A, __m512d __B)
@@ -974,18 +978,18 @@ _mm512_maskz_max_pd (__mmask8 __U, __m512d __A, __m512d __B)
974
978
  }
975
979
 
976
980
  #define _mm512_max_round_ps(A, B, R) \
977
- (__m512)__builtin_ia32_maxps512((__v16sf)(__m512)(A), \
978
- (__v16sf)(__m512)(B), (int)(R))
981
+ ((__m512)__builtin_ia32_maxps512((__v16sf)(__m512)(A), \
982
+ (__v16sf)(__m512)(B), (int)(R)))
979
983
 
980
984
  #define _mm512_mask_max_round_ps(W, U, A, B, R) \
981
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
985
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
982
986
  (__v16sf)_mm512_max_round_ps((A), (B), (R)), \
983
- (__v16sf)(W))
987
+ (__v16sf)(W)))
984
988
 
985
989
  #define _mm512_maskz_max_round_ps(U, A, B, R) \
986
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
990
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
987
991
  (__v16sf)_mm512_max_round_ps((A), (B), (R)), \
988
- (__v16sf)_mm512_setzero_ps())
992
+ (__v16sf)_mm512_setzero_ps()))
989
993
 
990
994
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
991
995
  _mm512_max_ps(__m512 __A, __m512 __B)
@@ -1029,22 +1033,22 @@ _mm_maskz_max_ss(__mmask8 __U,__m128 __A, __m128 __B) {
1029
1033
  }
1030
1034
 
1031
1035
  #define _mm_max_round_ss(A, B, R) \
1032
- (__m128)__builtin_ia32_maxss_round_mask((__v4sf)(__m128)(A), \
1033
- (__v4sf)(__m128)(B), \
1034
- (__v4sf)_mm_setzero_ps(), \
1035
- (__mmask8)-1, (int)(R))
1036
+ ((__m128)__builtin_ia32_maxss_round_mask((__v4sf)(__m128)(A), \
1037
+ (__v4sf)(__m128)(B), \
1038
+ (__v4sf)_mm_setzero_ps(), \
1039
+ (__mmask8)-1, (int)(R)))
1036
1040
 
1037
1041
  #define _mm_mask_max_round_ss(W, U, A, B, R) \
1038
- (__m128)__builtin_ia32_maxss_round_mask((__v4sf)(__m128)(A), \
1039
- (__v4sf)(__m128)(B), \
1040
- (__v4sf)(__m128)(W), (__mmask8)(U), \
1041
- (int)(R))
1042
+ ((__m128)__builtin_ia32_maxss_round_mask((__v4sf)(__m128)(A), \
1043
+ (__v4sf)(__m128)(B), \
1044
+ (__v4sf)(__m128)(W), (__mmask8)(U), \
1045
+ (int)(R)))
1042
1046
 
1043
1047
  #define _mm_maskz_max_round_ss(U, A, B, R) \
1044
- (__m128)__builtin_ia32_maxss_round_mask((__v4sf)(__m128)(A), \
1045
- (__v4sf)(__m128)(B), \
1046
- (__v4sf)_mm_setzero_ps(), \
1047
- (__mmask8)(U), (int)(R))
1048
+ ((__m128)__builtin_ia32_maxss_round_mask((__v4sf)(__m128)(A), \
1049
+ (__v4sf)(__m128)(B), \
1050
+ (__v4sf)_mm_setzero_ps(), \
1051
+ (__mmask8)(U), (int)(R)))
1048
1052
 
1049
1053
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
1050
1054
  _mm_mask_max_sd(__m128d __W, __mmask8 __U,__m128d __A, __m128d __B) {
@@ -1065,28 +1069,28 @@ _mm_maskz_max_sd(__mmask8 __U,__m128d __A, __m128d __B) {
1065
1069
  }
1066
1070
 
1067
1071
  #define _mm_max_round_sd(A, B, R) \
1068
- (__m128d)__builtin_ia32_maxsd_round_mask((__v2df)(__m128d)(A), \
1069
- (__v2df)(__m128d)(B), \
1070
- (__v2df)_mm_setzero_pd(), \
1071
- (__mmask8)-1, (int)(R))
1072
+ ((__m128d)__builtin_ia32_maxsd_round_mask((__v2df)(__m128d)(A), \
1073
+ (__v2df)(__m128d)(B), \
1074
+ (__v2df)_mm_setzero_pd(), \
1075
+ (__mmask8)-1, (int)(R)))
1072
1076
 
1073
1077
  #define _mm_mask_max_round_sd(W, U, A, B, R) \
1074
- (__m128d)__builtin_ia32_maxsd_round_mask((__v2df)(__m128d)(A), \
1075
- (__v2df)(__m128d)(B), \
1076
- (__v2df)(__m128d)(W), \
1077
- (__mmask8)(U), (int)(R))
1078
+ ((__m128d)__builtin_ia32_maxsd_round_mask((__v2df)(__m128d)(A), \
1079
+ (__v2df)(__m128d)(B), \
1080
+ (__v2df)(__m128d)(W), \
1081
+ (__mmask8)(U), (int)(R)))
1078
1082
 
1079
1083
  #define _mm_maskz_max_round_sd(U, A, B, R) \
1080
- (__m128d)__builtin_ia32_maxsd_round_mask((__v2df)(__m128d)(A), \
1081
- (__v2df)(__m128d)(B), \
1082
- (__v2df)_mm_setzero_pd(), \
1083
- (__mmask8)(U), (int)(R))
1084
+ ((__m128d)__builtin_ia32_maxsd_round_mask((__v2df)(__m128d)(A), \
1085
+ (__v2df)(__m128d)(B), \
1086
+ (__v2df)_mm_setzero_pd(), \
1087
+ (__mmask8)(U), (int)(R)))
1084
1088
 
1085
1089
  static __inline __m512i
1086
1090
  __DEFAULT_FN_ATTRS512
1087
1091
  _mm512_max_epi32(__m512i __A, __m512i __B)
1088
1092
  {
1089
- return (__m512i)__builtin_ia32_pmaxsd512((__v16si)__A, (__v16si)__B);
1093
+ return (__m512i)__builtin_elementwise_max((__v16si)__A, (__v16si)__B);
1090
1094
  }
1091
1095
 
1092
1096
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
@@ -1108,7 +1112,7 @@ _mm512_maskz_max_epi32 (__mmask16 __M, __m512i __A, __m512i __B)
1108
1112
  static __inline __m512i __DEFAULT_FN_ATTRS512
1109
1113
  _mm512_max_epu32(__m512i __A, __m512i __B)
1110
1114
  {
1111
- return (__m512i)__builtin_ia32_pmaxud512((__v16si)__A, (__v16si)__B);
1115
+ return (__m512i)__builtin_elementwise_max((__v16su)__A, (__v16su)__B);
1112
1116
  }
1113
1117
 
1114
1118
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
@@ -1130,7 +1134,7 @@ _mm512_maskz_max_epu32 (__mmask16 __M, __m512i __A, __m512i __B)
1130
1134
  static __inline __m512i __DEFAULT_FN_ATTRS512
1131
1135
  _mm512_max_epi64(__m512i __A, __m512i __B)
1132
1136
  {
1133
- return (__m512i)__builtin_ia32_pmaxsq512((__v8di)__A, (__v8di)__B);
1137
+ return (__m512i)__builtin_elementwise_max((__v8di)__A, (__v8di)__B);
1134
1138
  }
1135
1139
 
1136
1140
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
@@ -1152,7 +1156,7 @@ _mm512_maskz_max_epi64 (__mmask8 __M, __m512i __A, __m512i __B)
1152
1156
  static __inline __m512i __DEFAULT_FN_ATTRS512
1153
1157
  _mm512_max_epu64(__m512i __A, __m512i __B)
1154
1158
  {
1155
- return (__m512i)__builtin_ia32_pmaxuq512((__v8di)__A, (__v8di)__B);
1159
+ return (__m512i)__builtin_elementwise_max((__v8du)__A, (__v8du)__B);
1156
1160
  }
1157
1161
 
1158
1162
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
@@ -1172,18 +1176,18 @@ _mm512_maskz_max_epu64 (__mmask8 __M, __m512i __A, __m512i __B)
1172
1176
  }
1173
1177
 
1174
1178
  #define _mm512_min_round_pd(A, B, R) \
1175
- (__m512d)__builtin_ia32_minpd512((__v8df)(__m512d)(A), \
1176
- (__v8df)(__m512d)(B), (int)(R))
1179
+ ((__m512d)__builtin_ia32_minpd512((__v8df)(__m512d)(A), \
1180
+ (__v8df)(__m512d)(B), (int)(R)))
1177
1181
 
1178
1182
  #define _mm512_mask_min_round_pd(W, U, A, B, R) \
1179
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
1183
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
1180
1184
  (__v8df)_mm512_min_round_pd((A), (B), (R)), \
1181
- (__v8df)(W))
1185
+ (__v8df)(W)))
1182
1186
 
1183
1187
  #define _mm512_maskz_min_round_pd(U, A, B, R) \
1184
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
1188
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
1185
1189
  (__v8df)_mm512_min_round_pd((A), (B), (R)), \
1186
- (__v8df)_mm512_setzero_pd())
1190
+ (__v8df)_mm512_setzero_pd()))
1187
1191
 
1188
1192
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
1189
1193
  _mm512_min_pd(__m512d __A, __m512d __B)
@@ -1209,18 +1213,18 @@ _mm512_maskz_min_pd (__mmask8 __U, __m512d __A, __m512d __B)
1209
1213
  }
1210
1214
 
1211
1215
  #define _mm512_min_round_ps(A, B, R) \
1212
- (__m512)__builtin_ia32_minps512((__v16sf)(__m512)(A), \
1213
- (__v16sf)(__m512)(B), (int)(R))
1216
+ ((__m512)__builtin_ia32_minps512((__v16sf)(__m512)(A), \
1217
+ (__v16sf)(__m512)(B), (int)(R)))
1214
1218
 
1215
1219
  #define _mm512_mask_min_round_ps(W, U, A, B, R) \
1216
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
1220
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
1217
1221
  (__v16sf)_mm512_min_round_ps((A), (B), (R)), \
1218
- (__v16sf)(W))
1222
+ (__v16sf)(W)))
1219
1223
 
1220
1224
  #define _mm512_maskz_min_round_ps(U, A, B, R) \
1221
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
1225
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
1222
1226
  (__v16sf)_mm512_min_round_ps((A), (B), (R)), \
1223
- (__v16sf)_mm512_setzero_ps())
1227
+ (__v16sf)_mm512_setzero_ps()))
1224
1228
 
1225
1229
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
1226
1230
  _mm512_min_ps(__m512 __A, __m512 __B)
@@ -1264,22 +1268,22 @@ _mm_maskz_min_ss(__mmask8 __U,__m128 __A, __m128 __B) {
1264
1268
  }
1265
1269
 
1266
1270
  #define _mm_min_round_ss(A, B, R) \
1267
- (__m128)__builtin_ia32_minss_round_mask((__v4sf)(__m128)(A), \
1268
- (__v4sf)(__m128)(B), \
1269
- (__v4sf)_mm_setzero_ps(), \
1270
- (__mmask8)-1, (int)(R))
1271
+ ((__m128)__builtin_ia32_minss_round_mask((__v4sf)(__m128)(A), \
1272
+ (__v4sf)(__m128)(B), \
1273
+ (__v4sf)_mm_setzero_ps(), \
1274
+ (__mmask8)-1, (int)(R)))
1271
1275
 
1272
1276
  #define _mm_mask_min_round_ss(W, U, A, B, R) \
1273
- (__m128)__builtin_ia32_minss_round_mask((__v4sf)(__m128)(A), \
1274
- (__v4sf)(__m128)(B), \
1275
- (__v4sf)(__m128)(W), (__mmask8)(U), \
1276
- (int)(R))
1277
+ ((__m128)__builtin_ia32_minss_round_mask((__v4sf)(__m128)(A), \
1278
+ (__v4sf)(__m128)(B), \
1279
+ (__v4sf)(__m128)(W), (__mmask8)(U), \
1280
+ (int)(R)))
1277
1281
 
1278
1282
  #define _mm_maskz_min_round_ss(U, A, B, R) \
1279
- (__m128)__builtin_ia32_minss_round_mask((__v4sf)(__m128)(A), \
1280
- (__v4sf)(__m128)(B), \
1281
- (__v4sf)_mm_setzero_ps(), \
1282
- (__mmask8)(U), (int)(R))
1283
+ ((__m128)__builtin_ia32_minss_round_mask((__v4sf)(__m128)(A), \
1284
+ (__v4sf)(__m128)(B), \
1285
+ (__v4sf)_mm_setzero_ps(), \
1286
+ (__mmask8)(U), (int)(R)))
1283
1287
 
1284
1288
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
1285
1289
  _mm_mask_min_sd(__m128d __W, __mmask8 __U,__m128d __A, __m128d __B) {
@@ -1300,28 +1304,28 @@ _mm_maskz_min_sd(__mmask8 __U,__m128d __A, __m128d __B) {
1300
1304
  }
1301
1305
 
1302
1306
  #define _mm_min_round_sd(A, B, R) \
1303
- (__m128d)__builtin_ia32_minsd_round_mask((__v2df)(__m128d)(A), \
1304
- (__v2df)(__m128d)(B), \
1305
- (__v2df)_mm_setzero_pd(), \
1306
- (__mmask8)-1, (int)(R))
1307
+ ((__m128d)__builtin_ia32_minsd_round_mask((__v2df)(__m128d)(A), \
1308
+ (__v2df)(__m128d)(B), \
1309
+ (__v2df)_mm_setzero_pd(), \
1310
+ (__mmask8)-1, (int)(R)))
1307
1311
 
1308
1312
  #define _mm_mask_min_round_sd(W, U, A, B, R) \
1309
- (__m128d)__builtin_ia32_minsd_round_mask((__v2df)(__m128d)(A), \
1310
- (__v2df)(__m128d)(B), \
1311
- (__v2df)(__m128d)(W), \
1312
- (__mmask8)(U), (int)(R))
1313
+ ((__m128d)__builtin_ia32_minsd_round_mask((__v2df)(__m128d)(A), \
1314
+ (__v2df)(__m128d)(B), \
1315
+ (__v2df)(__m128d)(W), \
1316
+ (__mmask8)(U), (int)(R)))
1313
1317
 
1314
1318
  #define _mm_maskz_min_round_sd(U, A, B, R) \
1315
- (__m128d)__builtin_ia32_minsd_round_mask((__v2df)(__m128d)(A), \
1316
- (__v2df)(__m128d)(B), \
1317
- (__v2df)_mm_setzero_pd(), \
1318
- (__mmask8)(U), (int)(R))
1319
+ ((__m128d)__builtin_ia32_minsd_round_mask((__v2df)(__m128d)(A), \
1320
+ (__v2df)(__m128d)(B), \
1321
+ (__v2df)_mm_setzero_pd(), \
1322
+ (__mmask8)(U), (int)(R)))
1319
1323
 
1320
1324
  static __inline __m512i
1321
1325
  __DEFAULT_FN_ATTRS512
1322
1326
  _mm512_min_epi32(__m512i __A, __m512i __B)
1323
1327
  {
1324
- return (__m512i)__builtin_ia32_pminsd512((__v16si)__A, (__v16si)__B);
1328
+ return (__m512i)__builtin_elementwise_min((__v16si)__A, (__v16si)__B);
1325
1329
  }
1326
1330
 
1327
1331
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
@@ -1343,7 +1347,7 @@ _mm512_maskz_min_epi32 (__mmask16 __M, __m512i __A, __m512i __B)
1343
1347
  static __inline __m512i __DEFAULT_FN_ATTRS512
1344
1348
  _mm512_min_epu32(__m512i __A, __m512i __B)
1345
1349
  {
1346
- return (__m512i)__builtin_ia32_pminud512((__v16si)__A, (__v16si)__B);
1350
+ return (__m512i)__builtin_elementwise_min((__v16su)__A, (__v16su)__B);
1347
1351
  }
1348
1352
 
1349
1353
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
@@ -1365,7 +1369,7 @@ _mm512_maskz_min_epu32 (__mmask16 __M, __m512i __A, __m512i __B)
1365
1369
  static __inline __m512i __DEFAULT_FN_ATTRS512
1366
1370
  _mm512_min_epi64(__m512i __A, __m512i __B)
1367
1371
  {
1368
- return (__m512i)__builtin_ia32_pminsq512((__v8di)__A, (__v8di)__B);
1372
+ return (__m512i)__builtin_elementwise_min((__v8di)__A, (__v8di)__B);
1369
1373
  }
1370
1374
 
1371
1375
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
@@ -1387,7 +1391,7 @@ _mm512_maskz_min_epi64 (__mmask8 __M, __m512i __A, __m512i __B)
1387
1391
  static __inline __m512i __DEFAULT_FN_ATTRS512
1388
1392
  _mm512_min_epu64(__m512i __A, __m512i __B)
1389
1393
  {
1390
- return (__m512i)__builtin_ia32_pminuq512((__v8di)__A, (__v8di)__B);
1394
+ return (__m512i)__builtin_elementwise_min((__v8du)__A, (__v8du)__B);
1391
1395
  }
1392
1396
 
1393
1397
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
@@ -1485,17 +1489,17 @@ _mm512_mask_mullox_epi64(__m512i __W, __mmask8 __U, __m512i __A, __m512i __B) {
1485
1489
  }
1486
1490
 
1487
1491
  #define _mm512_sqrt_round_pd(A, R) \
1488
- (__m512d)__builtin_ia32_sqrtpd512((__v8df)(__m512d)(A), (int)(R))
1492
+ ((__m512d)__builtin_ia32_sqrtpd512((__v8df)(__m512d)(A), (int)(R)))
1489
1493
 
1490
1494
  #define _mm512_mask_sqrt_round_pd(W, U, A, R) \
1491
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
1495
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
1492
1496
  (__v8df)_mm512_sqrt_round_pd((A), (R)), \
1493
- (__v8df)(__m512d)(W))
1497
+ (__v8df)(__m512d)(W)))
1494
1498
 
1495
1499
  #define _mm512_maskz_sqrt_round_pd(U, A, R) \
1496
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
1500
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
1497
1501
  (__v8df)_mm512_sqrt_round_pd((A), (R)), \
1498
- (__v8df)_mm512_setzero_pd())
1502
+ (__v8df)_mm512_setzero_pd()))
1499
1503
 
1500
1504
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
1501
1505
  _mm512_sqrt_pd(__m512d __A)
@@ -1521,17 +1525,17 @@ _mm512_maskz_sqrt_pd (__mmask8 __U, __m512d __A)
1521
1525
  }
1522
1526
 
1523
1527
  #define _mm512_sqrt_round_ps(A, R) \
1524
- (__m512)__builtin_ia32_sqrtps512((__v16sf)(__m512)(A), (int)(R))
1528
+ ((__m512)__builtin_ia32_sqrtps512((__v16sf)(__m512)(A), (int)(R)))
1525
1529
 
1526
1530
  #define _mm512_mask_sqrt_round_ps(W, U, A, R) \
1527
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
1531
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
1528
1532
  (__v16sf)_mm512_sqrt_round_ps((A), (R)), \
1529
- (__v16sf)(__m512)(W))
1533
+ (__v16sf)(__m512)(W)))
1530
1534
 
1531
1535
  #define _mm512_maskz_sqrt_round_ps(U, A, R) \
1532
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
1536
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
1533
1537
  (__v16sf)_mm512_sqrt_round_ps((A), (R)), \
1534
- (__v16sf)_mm512_setzero_ps())
1538
+ (__v16sf)_mm512_setzero_ps()))
1535
1539
 
1536
1540
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
1537
1541
  _mm512_sqrt_ps(__m512 __A)
@@ -1846,7 +1850,7 @@ _mm512_mask_ceil_pd (__m512d __W, __mmask8 __U, __m512d __A)
1846
1850
  static __inline __m512i __DEFAULT_FN_ATTRS512
1847
1851
  _mm512_abs_epi64(__m512i __A)
1848
1852
  {
1849
- return (__m512i)__builtin_ia32_pabsq512((__v8di)__A);
1853
+ return (__m512i)__builtin_elementwise_abs((__v8di)__A);
1850
1854
  }
1851
1855
 
1852
1856
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
@@ -1868,7 +1872,7 @@ _mm512_maskz_abs_epi64 (__mmask8 __U, __m512i __A)
1868
1872
  static __inline __m512i __DEFAULT_FN_ATTRS512
1869
1873
  _mm512_abs_epi32(__m512i __A)
1870
1874
  {
1871
- return (__m512i)__builtin_ia32_pabsd512((__v16si) __A);
1875
+ return (__m512i)__builtin_elementwise_abs((__v16si) __A);
1872
1876
  }
1873
1877
 
1874
1878
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
@@ -1900,22 +1904,22 @@ _mm_maskz_add_ss(__mmask8 __U,__m128 __A, __m128 __B) {
1900
1904
  }
1901
1905
 
1902
1906
  #define _mm_add_round_ss(A, B, R) \
1903
- (__m128)__builtin_ia32_addss_round_mask((__v4sf)(__m128)(A), \
1904
- (__v4sf)(__m128)(B), \
1905
- (__v4sf)_mm_setzero_ps(), \
1906
- (__mmask8)-1, (int)(R))
1907
+ ((__m128)__builtin_ia32_addss_round_mask((__v4sf)(__m128)(A), \
1908
+ (__v4sf)(__m128)(B), \
1909
+ (__v4sf)_mm_setzero_ps(), \
1910
+ (__mmask8)-1, (int)(R)))
1907
1911
 
1908
1912
  #define _mm_mask_add_round_ss(W, U, A, B, R) \
1909
- (__m128)__builtin_ia32_addss_round_mask((__v4sf)(__m128)(A), \
1910
- (__v4sf)(__m128)(B), \
1911
- (__v4sf)(__m128)(W), (__mmask8)(U), \
1912
- (int)(R))
1913
+ ((__m128)__builtin_ia32_addss_round_mask((__v4sf)(__m128)(A), \
1914
+ (__v4sf)(__m128)(B), \
1915
+ (__v4sf)(__m128)(W), (__mmask8)(U), \
1916
+ (int)(R)))
1913
1917
 
1914
1918
  #define _mm_maskz_add_round_ss(U, A, B, R) \
1915
- (__m128)__builtin_ia32_addss_round_mask((__v4sf)(__m128)(A), \
1916
- (__v4sf)(__m128)(B), \
1917
- (__v4sf)_mm_setzero_ps(), \
1918
- (__mmask8)(U), (int)(R))
1919
+ ((__m128)__builtin_ia32_addss_round_mask((__v4sf)(__m128)(A), \
1920
+ (__v4sf)(__m128)(B), \
1921
+ (__v4sf)_mm_setzero_ps(), \
1922
+ (__mmask8)(U), (int)(R)))
1919
1923
 
1920
1924
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
1921
1925
  _mm_mask_add_sd(__m128d __W, __mmask8 __U,__m128d __A, __m128d __B) {
@@ -1929,22 +1933,22 @@ _mm_maskz_add_sd(__mmask8 __U,__m128d __A, __m128d __B) {
1929
1933
  return __builtin_ia32_selectsd_128(__U, __A, _mm_setzero_pd());
1930
1934
  }
1931
1935
  #define _mm_add_round_sd(A, B, R) \
1932
- (__m128d)__builtin_ia32_addsd_round_mask((__v2df)(__m128d)(A), \
1933
- (__v2df)(__m128d)(B), \
1934
- (__v2df)_mm_setzero_pd(), \
1935
- (__mmask8)-1, (int)(R))
1936
+ ((__m128d)__builtin_ia32_addsd_round_mask((__v2df)(__m128d)(A), \
1937
+ (__v2df)(__m128d)(B), \
1938
+ (__v2df)_mm_setzero_pd(), \
1939
+ (__mmask8)-1, (int)(R)))
1936
1940
 
1937
1941
  #define _mm_mask_add_round_sd(W, U, A, B, R) \
1938
- (__m128d)__builtin_ia32_addsd_round_mask((__v2df)(__m128d)(A), \
1939
- (__v2df)(__m128d)(B), \
1940
- (__v2df)(__m128d)(W), \
1941
- (__mmask8)(U), (int)(R))
1942
+ ((__m128d)__builtin_ia32_addsd_round_mask((__v2df)(__m128d)(A), \
1943
+ (__v2df)(__m128d)(B), \
1944
+ (__v2df)(__m128d)(W), \
1945
+ (__mmask8)(U), (int)(R)))
1942
1946
 
1943
1947
  #define _mm_maskz_add_round_sd(U, A, B, R) \
1944
- (__m128d)__builtin_ia32_addsd_round_mask((__v2df)(__m128d)(A), \
1945
- (__v2df)(__m128d)(B), \
1946
- (__v2df)_mm_setzero_pd(), \
1947
- (__mmask8)(U), (int)(R))
1948
+ ((__m128d)__builtin_ia32_addsd_round_mask((__v2df)(__m128d)(A), \
1949
+ (__v2df)(__m128d)(B), \
1950
+ (__v2df)_mm_setzero_pd(), \
1951
+ (__mmask8)(U), (int)(R)))
1948
1952
 
1949
1953
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
1950
1954
  _mm512_mask_add_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512d __B) {
@@ -1975,32 +1979,32 @@ _mm512_maskz_add_ps(__mmask16 __U, __m512 __A, __m512 __B) {
1975
1979
  }
1976
1980
 
1977
1981
  #define _mm512_add_round_pd(A, B, R) \
1978
- (__m512d)__builtin_ia32_addpd512((__v8df)(__m512d)(A), \
1979
- (__v8df)(__m512d)(B), (int)(R))
1982
+ ((__m512d)__builtin_ia32_addpd512((__v8df)(__m512d)(A), \
1983
+ (__v8df)(__m512d)(B), (int)(R)))
1980
1984
 
1981
1985
  #define _mm512_mask_add_round_pd(W, U, A, B, R) \
1982
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
1986
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
1983
1987
  (__v8df)_mm512_add_round_pd((A), (B), (R)), \
1984
- (__v8df)(__m512d)(W))
1988
+ (__v8df)(__m512d)(W)))
1985
1989
 
1986
1990
  #define _mm512_maskz_add_round_pd(U, A, B, R) \
1987
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
1991
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
1988
1992
  (__v8df)_mm512_add_round_pd((A), (B), (R)), \
1989
- (__v8df)_mm512_setzero_pd())
1993
+ (__v8df)_mm512_setzero_pd()))
1990
1994
 
1991
1995
  #define _mm512_add_round_ps(A, B, R) \
1992
- (__m512)__builtin_ia32_addps512((__v16sf)(__m512)(A), \
1993
- (__v16sf)(__m512)(B), (int)(R))
1996
+ ((__m512)__builtin_ia32_addps512((__v16sf)(__m512)(A), \
1997
+ (__v16sf)(__m512)(B), (int)(R)))
1994
1998
 
1995
1999
  #define _mm512_mask_add_round_ps(W, U, A, B, R) \
1996
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2000
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
1997
2001
  (__v16sf)_mm512_add_round_ps((A), (B), (R)), \
1998
- (__v16sf)(__m512)(W))
2002
+ (__v16sf)(__m512)(W)))
1999
2003
 
2000
2004
  #define _mm512_maskz_add_round_ps(U, A, B, R) \
2001
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2005
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2002
2006
  (__v16sf)_mm512_add_round_ps((A), (B), (R)), \
2003
- (__v16sf)_mm512_setzero_ps())
2007
+ (__v16sf)_mm512_setzero_ps()))
2004
2008
 
2005
2009
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
2006
2010
  _mm_mask_sub_ss(__m128 __W, __mmask8 __U,__m128 __A, __m128 __B) {
@@ -2014,22 +2018,22 @@ _mm_maskz_sub_ss(__mmask8 __U,__m128 __A, __m128 __B) {
2014
2018
  return __builtin_ia32_selectss_128(__U, __A, _mm_setzero_ps());
2015
2019
  }
2016
2020
  #define _mm_sub_round_ss(A, B, R) \
2017
- (__m128)__builtin_ia32_subss_round_mask((__v4sf)(__m128)(A), \
2018
- (__v4sf)(__m128)(B), \
2019
- (__v4sf)_mm_setzero_ps(), \
2020
- (__mmask8)-1, (int)(R))
2021
+ ((__m128)__builtin_ia32_subss_round_mask((__v4sf)(__m128)(A), \
2022
+ (__v4sf)(__m128)(B), \
2023
+ (__v4sf)_mm_setzero_ps(), \
2024
+ (__mmask8)-1, (int)(R)))
2021
2025
 
2022
2026
  #define _mm_mask_sub_round_ss(W, U, A, B, R) \
2023
- (__m128)__builtin_ia32_subss_round_mask((__v4sf)(__m128)(A), \
2024
- (__v4sf)(__m128)(B), \
2025
- (__v4sf)(__m128)(W), (__mmask8)(U), \
2026
- (int)(R))
2027
+ ((__m128)__builtin_ia32_subss_round_mask((__v4sf)(__m128)(A), \
2028
+ (__v4sf)(__m128)(B), \
2029
+ (__v4sf)(__m128)(W), (__mmask8)(U), \
2030
+ (int)(R)))
2027
2031
 
2028
2032
  #define _mm_maskz_sub_round_ss(U, A, B, R) \
2029
- (__m128)__builtin_ia32_subss_round_mask((__v4sf)(__m128)(A), \
2030
- (__v4sf)(__m128)(B), \
2031
- (__v4sf)_mm_setzero_ps(), \
2032
- (__mmask8)(U), (int)(R))
2033
+ ((__m128)__builtin_ia32_subss_round_mask((__v4sf)(__m128)(A), \
2034
+ (__v4sf)(__m128)(B), \
2035
+ (__v4sf)_mm_setzero_ps(), \
2036
+ (__mmask8)(U), (int)(R)))
2033
2037
 
2034
2038
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
2035
2039
  _mm_mask_sub_sd(__m128d __W, __mmask8 __U,__m128d __A, __m128d __B) {
@@ -2044,22 +2048,22 @@ _mm_maskz_sub_sd(__mmask8 __U,__m128d __A, __m128d __B) {
2044
2048
  }
2045
2049
 
2046
2050
  #define _mm_sub_round_sd(A, B, R) \
2047
- (__m128d)__builtin_ia32_subsd_round_mask((__v2df)(__m128d)(A), \
2048
- (__v2df)(__m128d)(B), \
2049
- (__v2df)_mm_setzero_pd(), \
2050
- (__mmask8)-1, (int)(R))
2051
+ ((__m128d)__builtin_ia32_subsd_round_mask((__v2df)(__m128d)(A), \
2052
+ (__v2df)(__m128d)(B), \
2053
+ (__v2df)_mm_setzero_pd(), \
2054
+ (__mmask8)-1, (int)(R)))
2051
2055
 
2052
2056
  #define _mm_mask_sub_round_sd(W, U, A, B, R) \
2053
- (__m128d)__builtin_ia32_subsd_round_mask((__v2df)(__m128d)(A), \
2054
- (__v2df)(__m128d)(B), \
2055
- (__v2df)(__m128d)(W), \
2056
- (__mmask8)(U), (int)(R))
2057
+ ((__m128d)__builtin_ia32_subsd_round_mask((__v2df)(__m128d)(A), \
2058
+ (__v2df)(__m128d)(B), \
2059
+ (__v2df)(__m128d)(W), \
2060
+ (__mmask8)(U), (int)(R)))
2057
2061
 
2058
2062
  #define _mm_maskz_sub_round_sd(U, A, B, R) \
2059
- (__m128d)__builtin_ia32_subsd_round_mask((__v2df)(__m128d)(A), \
2060
- (__v2df)(__m128d)(B), \
2061
- (__v2df)_mm_setzero_pd(), \
2062
- (__mmask8)(U), (int)(R))
2063
+ ((__m128d)__builtin_ia32_subsd_round_mask((__v2df)(__m128d)(A), \
2064
+ (__v2df)(__m128d)(B), \
2065
+ (__v2df)_mm_setzero_pd(), \
2066
+ (__mmask8)(U), (int)(R)))
2063
2067
 
2064
2068
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
2065
2069
  _mm512_mask_sub_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512d __B) {
@@ -2090,32 +2094,32 @@ _mm512_maskz_sub_ps(__mmask16 __U, __m512 __A, __m512 __B) {
2090
2094
  }
2091
2095
 
2092
2096
  #define _mm512_sub_round_pd(A, B, R) \
2093
- (__m512d)__builtin_ia32_subpd512((__v8df)(__m512d)(A), \
2094
- (__v8df)(__m512d)(B), (int)(R))
2097
+ ((__m512d)__builtin_ia32_subpd512((__v8df)(__m512d)(A), \
2098
+ (__v8df)(__m512d)(B), (int)(R)))
2095
2099
 
2096
2100
  #define _mm512_mask_sub_round_pd(W, U, A, B, R) \
2097
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
2101
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
2098
2102
  (__v8df)_mm512_sub_round_pd((A), (B), (R)), \
2099
- (__v8df)(__m512d)(W))
2103
+ (__v8df)(__m512d)(W)))
2100
2104
 
2101
2105
  #define _mm512_maskz_sub_round_pd(U, A, B, R) \
2102
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
2106
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
2103
2107
  (__v8df)_mm512_sub_round_pd((A), (B), (R)), \
2104
- (__v8df)_mm512_setzero_pd())
2108
+ (__v8df)_mm512_setzero_pd()))
2105
2109
 
2106
2110
  #define _mm512_sub_round_ps(A, B, R) \
2107
- (__m512)__builtin_ia32_subps512((__v16sf)(__m512)(A), \
2108
- (__v16sf)(__m512)(B), (int)(R))
2111
+ ((__m512)__builtin_ia32_subps512((__v16sf)(__m512)(A), \
2112
+ (__v16sf)(__m512)(B), (int)(R)))
2109
2113
 
2110
2114
  #define _mm512_mask_sub_round_ps(W, U, A, B, R) \
2111
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2115
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2112
2116
  (__v16sf)_mm512_sub_round_ps((A), (B), (R)), \
2113
- (__v16sf)(__m512)(W))
2117
+ (__v16sf)(__m512)(W)))
2114
2118
 
2115
2119
  #define _mm512_maskz_sub_round_ps(U, A, B, R) \
2116
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2120
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2117
2121
  (__v16sf)_mm512_sub_round_ps((A), (B), (R)), \
2118
- (__v16sf)_mm512_setzero_ps())
2122
+ (__v16sf)_mm512_setzero_ps()))
2119
2123
 
2120
2124
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
2121
2125
  _mm_mask_mul_ss(__m128 __W, __mmask8 __U,__m128 __A, __m128 __B) {
@@ -2129,22 +2133,22 @@ _mm_maskz_mul_ss(__mmask8 __U,__m128 __A, __m128 __B) {
2129
2133
  return __builtin_ia32_selectss_128(__U, __A, _mm_setzero_ps());
2130
2134
  }
2131
2135
  #define _mm_mul_round_ss(A, B, R) \
2132
- (__m128)__builtin_ia32_mulss_round_mask((__v4sf)(__m128)(A), \
2133
- (__v4sf)(__m128)(B), \
2134
- (__v4sf)_mm_setzero_ps(), \
2135
- (__mmask8)-1, (int)(R))
2136
+ ((__m128)__builtin_ia32_mulss_round_mask((__v4sf)(__m128)(A), \
2137
+ (__v4sf)(__m128)(B), \
2138
+ (__v4sf)_mm_setzero_ps(), \
2139
+ (__mmask8)-1, (int)(R)))
2136
2140
 
2137
2141
  #define _mm_mask_mul_round_ss(W, U, A, B, R) \
2138
- (__m128)__builtin_ia32_mulss_round_mask((__v4sf)(__m128)(A), \
2139
- (__v4sf)(__m128)(B), \
2140
- (__v4sf)(__m128)(W), (__mmask8)(U), \
2141
- (int)(R))
2142
+ ((__m128)__builtin_ia32_mulss_round_mask((__v4sf)(__m128)(A), \
2143
+ (__v4sf)(__m128)(B), \
2144
+ (__v4sf)(__m128)(W), (__mmask8)(U), \
2145
+ (int)(R)))
2142
2146
 
2143
2147
  #define _mm_maskz_mul_round_ss(U, A, B, R) \
2144
- (__m128)__builtin_ia32_mulss_round_mask((__v4sf)(__m128)(A), \
2145
- (__v4sf)(__m128)(B), \
2146
- (__v4sf)_mm_setzero_ps(), \
2147
- (__mmask8)(U), (int)(R))
2148
+ ((__m128)__builtin_ia32_mulss_round_mask((__v4sf)(__m128)(A), \
2149
+ (__v4sf)(__m128)(B), \
2150
+ (__v4sf)_mm_setzero_ps(), \
2151
+ (__mmask8)(U), (int)(R)))
2148
2152
 
2149
2153
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
2150
2154
  _mm_mask_mul_sd(__m128d __W, __mmask8 __U,__m128d __A, __m128d __B) {
@@ -2159,22 +2163,22 @@ _mm_maskz_mul_sd(__mmask8 __U,__m128d __A, __m128d __B) {
2159
2163
  }
2160
2164
 
2161
2165
  #define _mm_mul_round_sd(A, B, R) \
2162
- (__m128d)__builtin_ia32_mulsd_round_mask((__v2df)(__m128d)(A), \
2163
- (__v2df)(__m128d)(B), \
2164
- (__v2df)_mm_setzero_pd(), \
2165
- (__mmask8)-1, (int)(R))
2166
+ ((__m128d)__builtin_ia32_mulsd_round_mask((__v2df)(__m128d)(A), \
2167
+ (__v2df)(__m128d)(B), \
2168
+ (__v2df)_mm_setzero_pd(), \
2169
+ (__mmask8)-1, (int)(R)))
2166
2170
 
2167
2171
  #define _mm_mask_mul_round_sd(W, U, A, B, R) \
2168
- (__m128d)__builtin_ia32_mulsd_round_mask((__v2df)(__m128d)(A), \
2169
- (__v2df)(__m128d)(B), \
2170
- (__v2df)(__m128d)(W), \
2171
- (__mmask8)(U), (int)(R))
2172
+ ((__m128d)__builtin_ia32_mulsd_round_mask((__v2df)(__m128d)(A), \
2173
+ (__v2df)(__m128d)(B), \
2174
+ (__v2df)(__m128d)(W), \
2175
+ (__mmask8)(U), (int)(R)))
2172
2176
 
2173
2177
  #define _mm_maskz_mul_round_sd(U, A, B, R) \
2174
- (__m128d)__builtin_ia32_mulsd_round_mask((__v2df)(__m128d)(A), \
2175
- (__v2df)(__m128d)(B), \
2176
- (__v2df)_mm_setzero_pd(), \
2177
- (__mmask8)(U), (int)(R))
2178
+ ((__m128d)__builtin_ia32_mulsd_round_mask((__v2df)(__m128d)(A), \
2179
+ (__v2df)(__m128d)(B), \
2180
+ (__v2df)_mm_setzero_pd(), \
2181
+ (__mmask8)(U), (int)(R)))
2178
2182
 
2179
2183
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
2180
2184
  _mm512_mask_mul_pd(__m512d __W, __mmask8 __U, __m512d __A, __m512d __B) {
@@ -2205,32 +2209,32 @@ _mm512_maskz_mul_ps(__mmask16 __U, __m512 __A, __m512 __B) {
2205
2209
  }
2206
2210
 
2207
2211
  #define _mm512_mul_round_pd(A, B, R) \
2208
- (__m512d)__builtin_ia32_mulpd512((__v8df)(__m512d)(A), \
2209
- (__v8df)(__m512d)(B), (int)(R))
2212
+ ((__m512d)__builtin_ia32_mulpd512((__v8df)(__m512d)(A), \
2213
+ (__v8df)(__m512d)(B), (int)(R)))
2210
2214
 
2211
2215
  #define _mm512_mask_mul_round_pd(W, U, A, B, R) \
2212
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
2216
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
2213
2217
  (__v8df)_mm512_mul_round_pd((A), (B), (R)), \
2214
- (__v8df)(__m512d)(W))
2218
+ (__v8df)(__m512d)(W)))
2215
2219
 
2216
2220
  #define _mm512_maskz_mul_round_pd(U, A, B, R) \
2217
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
2221
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
2218
2222
  (__v8df)_mm512_mul_round_pd((A), (B), (R)), \
2219
- (__v8df)_mm512_setzero_pd())
2223
+ (__v8df)_mm512_setzero_pd()))
2220
2224
 
2221
2225
  #define _mm512_mul_round_ps(A, B, R) \
2222
- (__m512)__builtin_ia32_mulps512((__v16sf)(__m512)(A), \
2223
- (__v16sf)(__m512)(B), (int)(R))
2226
+ ((__m512)__builtin_ia32_mulps512((__v16sf)(__m512)(A), \
2227
+ (__v16sf)(__m512)(B), (int)(R)))
2224
2228
 
2225
2229
  #define _mm512_mask_mul_round_ps(W, U, A, B, R) \
2226
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2230
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2227
2231
  (__v16sf)_mm512_mul_round_ps((A), (B), (R)), \
2228
- (__v16sf)(__m512)(W))
2232
+ (__v16sf)(__m512)(W)))
2229
2233
 
2230
2234
  #define _mm512_maskz_mul_round_ps(U, A, B, R) \
2231
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2235
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2232
2236
  (__v16sf)_mm512_mul_round_ps((A), (B), (R)), \
2233
- (__v16sf)_mm512_setzero_ps())
2237
+ (__v16sf)_mm512_setzero_ps()))
2234
2238
 
2235
2239
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
2236
2240
  _mm_mask_div_ss(__m128 __W, __mmask8 __U,__m128 __A, __m128 __B) {
@@ -2245,22 +2249,22 @@ _mm_maskz_div_ss(__mmask8 __U,__m128 __A, __m128 __B) {
2245
2249
  }
2246
2250
 
2247
2251
  #define _mm_div_round_ss(A, B, R) \
2248
- (__m128)__builtin_ia32_divss_round_mask((__v4sf)(__m128)(A), \
2249
- (__v4sf)(__m128)(B), \
2250
- (__v4sf)_mm_setzero_ps(), \
2251
- (__mmask8)-1, (int)(R))
2252
+ ((__m128)__builtin_ia32_divss_round_mask((__v4sf)(__m128)(A), \
2253
+ (__v4sf)(__m128)(B), \
2254
+ (__v4sf)_mm_setzero_ps(), \
2255
+ (__mmask8)-1, (int)(R)))
2252
2256
 
2253
2257
  #define _mm_mask_div_round_ss(W, U, A, B, R) \
2254
- (__m128)__builtin_ia32_divss_round_mask((__v4sf)(__m128)(A), \
2255
- (__v4sf)(__m128)(B), \
2256
- (__v4sf)(__m128)(W), (__mmask8)(U), \
2257
- (int)(R))
2258
+ ((__m128)__builtin_ia32_divss_round_mask((__v4sf)(__m128)(A), \
2259
+ (__v4sf)(__m128)(B), \
2260
+ (__v4sf)(__m128)(W), (__mmask8)(U), \
2261
+ (int)(R)))
2258
2262
 
2259
2263
  #define _mm_maskz_div_round_ss(U, A, B, R) \
2260
- (__m128)__builtin_ia32_divss_round_mask((__v4sf)(__m128)(A), \
2261
- (__v4sf)(__m128)(B), \
2262
- (__v4sf)_mm_setzero_ps(), \
2263
- (__mmask8)(U), (int)(R))
2264
+ ((__m128)__builtin_ia32_divss_round_mask((__v4sf)(__m128)(A), \
2265
+ (__v4sf)(__m128)(B), \
2266
+ (__v4sf)_mm_setzero_ps(), \
2267
+ (__mmask8)(U), (int)(R)))
2264
2268
 
2265
2269
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
2266
2270
  _mm_mask_div_sd(__m128d __W, __mmask8 __U,__m128d __A, __m128d __B) {
@@ -2275,22 +2279,22 @@ _mm_maskz_div_sd(__mmask8 __U,__m128d __A, __m128d __B) {
2275
2279
  }
2276
2280
 
2277
2281
  #define _mm_div_round_sd(A, B, R) \
2278
- (__m128d)__builtin_ia32_divsd_round_mask((__v2df)(__m128d)(A), \
2279
- (__v2df)(__m128d)(B), \
2280
- (__v2df)_mm_setzero_pd(), \
2281
- (__mmask8)-1, (int)(R))
2282
+ ((__m128d)__builtin_ia32_divsd_round_mask((__v2df)(__m128d)(A), \
2283
+ (__v2df)(__m128d)(B), \
2284
+ (__v2df)_mm_setzero_pd(), \
2285
+ (__mmask8)-1, (int)(R)))
2282
2286
 
2283
2287
  #define _mm_mask_div_round_sd(W, U, A, B, R) \
2284
- (__m128d)__builtin_ia32_divsd_round_mask((__v2df)(__m128d)(A), \
2285
- (__v2df)(__m128d)(B), \
2286
- (__v2df)(__m128d)(W), \
2287
- (__mmask8)(U), (int)(R))
2288
+ ((__m128d)__builtin_ia32_divsd_round_mask((__v2df)(__m128d)(A), \
2289
+ (__v2df)(__m128d)(B), \
2290
+ (__v2df)(__m128d)(W), \
2291
+ (__mmask8)(U), (int)(R)))
2288
2292
 
2289
2293
  #define _mm_maskz_div_round_sd(U, A, B, R) \
2290
- (__m128d)__builtin_ia32_divsd_round_mask((__v2df)(__m128d)(A), \
2291
- (__v2df)(__m128d)(B), \
2292
- (__v2df)_mm_setzero_pd(), \
2293
- (__mmask8)(U), (int)(R))
2294
+ ((__m128d)__builtin_ia32_divsd_round_mask((__v2df)(__m128d)(A), \
2295
+ (__v2df)(__m128d)(B), \
2296
+ (__v2df)_mm_setzero_pd(), \
2297
+ (__mmask8)(U), (int)(R)))
2294
2298
 
2295
2299
  static __inline __m512d __DEFAULT_FN_ATTRS512
2296
2300
  _mm512_div_pd(__m512d __a, __m512d __b)
@@ -2333,179 +2337,179 @@ _mm512_maskz_div_ps(__mmask16 __U, __m512 __A, __m512 __B) {
2333
2337
  }
2334
2338
 
2335
2339
  #define _mm512_div_round_pd(A, B, R) \
2336
- (__m512d)__builtin_ia32_divpd512((__v8df)(__m512d)(A), \
2337
- (__v8df)(__m512d)(B), (int)(R))
2340
+ ((__m512d)__builtin_ia32_divpd512((__v8df)(__m512d)(A), \
2341
+ (__v8df)(__m512d)(B), (int)(R)))
2338
2342
 
2339
2343
  #define _mm512_mask_div_round_pd(W, U, A, B, R) \
2340
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
2344
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
2341
2345
  (__v8df)_mm512_div_round_pd((A), (B), (R)), \
2342
- (__v8df)(__m512d)(W))
2346
+ (__v8df)(__m512d)(W)))
2343
2347
 
2344
2348
  #define _mm512_maskz_div_round_pd(U, A, B, R) \
2345
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
2349
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
2346
2350
  (__v8df)_mm512_div_round_pd((A), (B), (R)), \
2347
- (__v8df)_mm512_setzero_pd())
2351
+ (__v8df)_mm512_setzero_pd()))
2348
2352
 
2349
2353
  #define _mm512_div_round_ps(A, B, R) \
2350
- (__m512)__builtin_ia32_divps512((__v16sf)(__m512)(A), \
2351
- (__v16sf)(__m512)(B), (int)(R))
2354
+ ((__m512)__builtin_ia32_divps512((__v16sf)(__m512)(A), \
2355
+ (__v16sf)(__m512)(B), (int)(R)))
2352
2356
 
2353
2357
  #define _mm512_mask_div_round_ps(W, U, A, B, R) \
2354
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2358
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2355
2359
  (__v16sf)_mm512_div_round_ps((A), (B), (R)), \
2356
- (__v16sf)(__m512)(W))
2360
+ (__v16sf)(__m512)(W)))
2357
2361
 
2358
2362
  #define _mm512_maskz_div_round_ps(U, A, B, R) \
2359
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2363
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
2360
2364
  (__v16sf)_mm512_div_round_ps((A), (B), (R)), \
2361
- (__v16sf)_mm512_setzero_ps())
2365
+ (__v16sf)_mm512_setzero_ps()))
2362
2366
 
2363
2367
  #define _mm512_roundscale_ps(A, B) \
2364
- (__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(A), (int)(B), \
2365
- (__v16sf)_mm512_undefined_ps(), \
2366
- (__mmask16)-1, \
2367
- _MM_FROUND_CUR_DIRECTION)
2368
+ ((__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(A), (int)(B), \
2369
+ (__v16sf)_mm512_undefined_ps(), \
2370
+ (__mmask16)-1, \
2371
+ _MM_FROUND_CUR_DIRECTION))
2368
2372
 
2369
2373
  #define _mm512_mask_roundscale_ps(A, B, C, imm) \
2370
- (__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(C), (int)(imm), \
2374
+ ((__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(C), (int)(imm), \
2371
2375
  (__v16sf)(__m512)(A), (__mmask16)(B), \
2372
- _MM_FROUND_CUR_DIRECTION)
2376
+ _MM_FROUND_CUR_DIRECTION))
2373
2377
 
2374
2378
  #define _mm512_maskz_roundscale_ps(A, B, imm) \
2375
- (__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(B), (int)(imm), \
2376
- (__v16sf)_mm512_setzero_ps(), \
2377
- (__mmask16)(A), \
2378
- _MM_FROUND_CUR_DIRECTION)
2379
+ ((__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(B), (int)(imm), \
2380
+ (__v16sf)_mm512_setzero_ps(), \
2381
+ (__mmask16)(A), \
2382
+ _MM_FROUND_CUR_DIRECTION))
2379
2383
 
2380
2384
  #define _mm512_mask_roundscale_round_ps(A, B, C, imm, R) \
2381
- (__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(C), (int)(imm), \
2385
+ ((__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(C), (int)(imm), \
2382
2386
  (__v16sf)(__m512)(A), (__mmask16)(B), \
2383
- (int)(R))
2387
+ (int)(R)))
2384
2388
 
2385
2389
  #define _mm512_maskz_roundscale_round_ps(A, B, imm, R) \
2386
- (__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(B), (int)(imm), \
2387
- (__v16sf)_mm512_setzero_ps(), \
2388
- (__mmask16)(A), (int)(R))
2390
+ ((__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(B), (int)(imm), \
2391
+ (__v16sf)_mm512_setzero_ps(), \
2392
+ (__mmask16)(A), (int)(R)))
2389
2393
 
2390
2394
  #define _mm512_roundscale_round_ps(A, imm, R) \
2391
- (__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(A), (int)(imm), \
2392
- (__v16sf)_mm512_undefined_ps(), \
2393
- (__mmask16)-1, (int)(R))
2395
+ ((__m512)__builtin_ia32_rndscaleps_mask((__v16sf)(__m512)(A), (int)(imm), \
2396
+ (__v16sf)_mm512_undefined_ps(), \
2397
+ (__mmask16)-1, (int)(R)))
2394
2398
 
2395
2399
  #define _mm512_roundscale_pd(A, B) \
2396
- (__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(A), (int)(B), \
2397
- (__v8df)_mm512_undefined_pd(), \
2398
- (__mmask8)-1, \
2399
- _MM_FROUND_CUR_DIRECTION)
2400
+ ((__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(A), (int)(B), \
2401
+ (__v8df)_mm512_undefined_pd(), \
2402
+ (__mmask8)-1, \
2403
+ _MM_FROUND_CUR_DIRECTION))
2400
2404
 
2401
2405
  #define _mm512_mask_roundscale_pd(A, B, C, imm) \
2402
- (__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(C), (int)(imm), \
2406
+ ((__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(C), (int)(imm), \
2403
2407
  (__v8df)(__m512d)(A), (__mmask8)(B), \
2404
- _MM_FROUND_CUR_DIRECTION)
2408
+ _MM_FROUND_CUR_DIRECTION))
2405
2409
 
2406
2410
  #define _mm512_maskz_roundscale_pd(A, B, imm) \
2407
- (__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(B), (int)(imm), \
2408
- (__v8df)_mm512_setzero_pd(), \
2409
- (__mmask8)(A), \
2410
- _MM_FROUND_CUR_DIRECTION)
2411
+ ((__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(B), (int)(imm), \
2412
+ (__v8df)_mm512_setzero_pd(), \
2413
+ (__mmask8)(A), \
2414
+ _MM_FROUND_CUR_DIRECTION))
2411
2415
 
2412
2416
  #define _mm512_mask_roundscale_round_pd(A, B, C, imm, R) \
2413
- (__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(C), (int)(imm), \
2417
+ ((__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(C), (int)(imm), \
2414
2418
  (__v8df)(__m512d)(A), (__mmask8)(B), \
2415
- (int)(R))
2419
+ (int)(R)))
2416
2420
 
2417
2421
  #define _mm512_maskz_roundscale_round_pd(A, B, imm, R) \
2418
- (__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(B), (int)(imm), \
2419
- (__v8df)_mm512_setzero_pd(), \
2420
- (__mmask8)(A), (int)(R))
2422
+ ((__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(B), (int)(imm), \
2423
+ (__v8df)_mm512_setzero_pd(), \
2424
+ (__mmask8)(A), (int)(R)))
2421
2425
 
2422
2426
  #define _mm512_roundscale_round_pd(A, imm, R) \
2423
- (__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(A), (int)(imm), \
2424
- (__v8df)_mm512_undefined_pd(), \
2425
- (__mmask8)-1, (int)(R))
2427
+ ((__m512d)__builtin_ia32_rndscalepd_mask((__v8df)(__m512d)(A), (int)(imm), \
2428
+ (__v8df)_mm512_undefined_pd(), \
2429
+ (__mmask8)-1, (int)(R)))
2426
2430
 
2427
2431
  #define _mm512_fmadd_round_pd(A, B, C, R) \
2428
- (__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
2429
- (__v8df)(__m512d)(B), \
2430
- (__v8df)(__m512d)(C), \
2431
- (__mmask8)-1, (int)(R))
2432
+ ((__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
2433
+ (__v8df)(__m512d)(B), \
2434
+ (__v8df)(__m512d)(C), \
2435
+ (__mmask8)-1, (int)(R)))
2432
2436
 
2433
2437
 
2434
2438
  #define _mm512_mask_fmadd_round_pd(A, U, B, C, R) \
2435
- (__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
2436
- (__v8df)(__m512d)(B), \
2437
- (__v8df)(__m512d)(C), \
2438
- (__mmask8)(U), (int)(R))
2439
+ ((__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
2440
+ (__v8df)(__m512d)(B), \
2441
+ (__v8df)(__m512d)(C), \
2442
+ (__mmask8)(U), (int)(R)))
2439
2443
 
2440
2444
 
2441
2445
  #define _mm512_mask3_fmadd_round_pd(A, B, C, U, R) \
2442
- (__m512d)__builtin_ia32_vfmaddpd512_mask3((__v8df)(__m512d)(A), \
2443
- (__v8df)(__m512d)(B), \
2444
- (__v8df)(__m512d)(C), \
2445
- (__mmask8)(U), (int)(R))
2446
+ ((__m512d)__builtin_ia32_vfmaddpd512_mask3((__v8df)(__m512d)(A), \
2447
+ (__v8df)(__m512d)(B), \
2448
+ (__v8df)(__m512d)(C), \
2449
+ (__mmask8)(U), (int)(R)))
2446
2450
 
2447
2451
 
2448
2452
  #define _mm512_maskz_fmadd_round_pd(U, A, B, C, R) \
2449
- (__m512d)__builtin_ia32_vfmaddpd512_maskz((__v8df)(__m512d)(A), \
2450
- (__v8df)(__m512d)(B), \
2451
- (__v8df)(__m512d)(C), \
2452
- (__mmask8)(U), (int)(R))
2453
+ ((__m512d)__builtin_ia32_vfmaddpd512_maskz((__v8df)(__m512d)(A), \
2454
+ (__v8df)(__m512d)(B), \
2455
+ (__v8df)(__m512d)(C), \
2456
+ (__mmask8)(U), (int)(R)))
2453
2457
 
2454
2458
 
2455
2459
  #define _mm512_fmsub_round_pd(A, B, C, R) \
2456
- (__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
2457
- (__v8df)(__m512d)(B), \
2458
- -(__v8df)(__m512d)(C), \
2459
- (__mmask8)-1, (int)(R))
2460
+ ((__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
2461
+ (__v8df)(__m512d)(B), \
2462
+ -(__v8df)(__m512d)(C), \
2463
+ (__mmask8)-1, (int)(R)))
2460
2464
 
2461
2465
 
2462
2466
  #define _mm512_mask_fmsub_round_pd(A, U, B, C, R) \
2463
- (__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
2464
- (__v8df)(__m512d)(B), \
2465
- -(__v8df)(__m512d)(C), \
2466
- (__mmask8)(U), (int)(R))
2467
+ ((__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
2468
+ (__v8df)(__m512d)(B), \
2469
+ -(__v8df)(__m512d)(C), \
2470
+ (__mmask8)(U), (int)(R)))
2467
2471
 
2468
2472
 
2469
2473
  #define _mm512_maskz_fmsub_round_pd(U, A, B, C, R) \
2470
- (__m512d)__builtin_ia32_vfmaddpd512_maskz((__v8df)(__m512d)(A), \
2471
- (__v8df)(__m512d)(B), \
2472
- -(__v8df)(__m512d)(C), \
2473
- (__mmask8)(U), (int)(R))
2474
+ ((__m512d)__builtin_ia32_vfmaddpd512_maskz((__v8df)(__m512d)(A), \
2475
+ (__v8df)(__m512d)(B), \
2476
+ -(__v8df)(__m512d)(C), \
2477
+ (__mmask8)(U), (int)(R)))
2474
2478
 
2475
2479
 
2476
2480
  #define _mm512_fnmadd_round_pd(A, B, C, R) \
2477
- (__m512d)__builtin_ia32_vfmaddpd512_mask(-(__v8df)(__m512d)(A), \
2478
- (__v8df)(__m512d)(B), \
2479
- (__v8df)(__m512d)(C), \
2480
- (__mmask8)-1, (int)(R))
2481
+ ((__m512d)__builtin_ia32_vfmaddpd512_mask(-(__v8df)(__m512d)(A), \
2482
+ (__v8df)(__m512d)(B), \
2483
+ (__v8df)(__m512d)(C), \
2484
+ (__mmask8)-1, (int)(R)))
2481
2485
 
2482
2486
 
2483
2487
  #define _mm512_mask3_fnmadd_round_pd(A, B, C, U, R) \
2484
- (__m512d)__builtin_ia32_vfmaddpd512_mask3(-(__v8df)(__m512d)(A), \
2485
- (__v8df)(__m512d)(B), \
2486
- (__v8df)(__m512d)(C), \
2487
- (__mmask8)(U), (int)(R))
2488
+ ((__m512d)__builtin_ia32_vfmaddpd512_mask3(-(__v8df)(__m512d)(A), \
2489
+ (__v8df)(__m512d)(B), \
2490
+ (__v8df)(__m512d)(C), \
2491
+ (__mmask8)(U), (int)(R)))
2488
2492
 
2489
2493
 
2490
2494
  #define _mm512_maskz_fnmadd_round_pd(U, A, B, C, R) \
2491
- (__m512d)__builtin_ia32_vfmaddpd512_maskz(-(__v8df)(__m512d)(A), \
2492
- (__v8df)(__m512d)(B), \
2493
- (__v8df)(__m512d)(C), \
2494
- (__mmask8)(U), (int)(R))
2495
+ ((__m512d)__builtin_ia32_vfmaddpd512_maskz(-(__v8df)(__m512d)(A), \
2496
+ (__v8df)(__m512d)(B), \
2497
+ (__v8df)(__m512d)(C), \
2498
+ (__mmask8)(U), (int)(R)))
2495
2499
 
2496
2500
 
2497
2501
  #define _mm512_fnmsub_round_pd(A, B, C, R) \
2498
- (__m512d)__builtin_ia32_vfmaddpd512_mask(-(__v8df)(__m512d)(A), \
2499
- (__v8df)(__m512d)(B), \
2500
- -(__v8df)(__m512d)(C), \
2501
- (__mmask8)-1, (int)(R))
2502
+ ((__m512d)__builtin_ia32_vfmaddpd512_mask(-(__v8df)(__m512d)(A), \
2503
+ (__v8df)(__m512d)(B), \
2504
+ -(__v8df)(__m512d)(C), \
2505
+ (__mmask8)-1, (int)(R)))
2502
2506
 
2503
2507
 
2504
2508
  #define _mm512_maskz_fnmsub_round_pd(U, A, B, C, R) \
2505
- (__m512d)__builtin_ia32_vfmaddpd512_maskz(-(__v8df)(__m512d)(A), \
2506
- (__v8df)(__m512d)(B), \
2507
- -(__v8df)(__m512d)(C), \
2508
- (__mmask8)(U), (int)(R))
2509
+ ((__m512d)__builtin_ia32_vfmaddpd512_maskz(-(__v8df)(__m512d)(A), \
2510
+ (__v8df)(__m512d)(B), \
2511
+ -(__v8df)(__m512d)(C), \
2512
+ (__mmask8)(U), (int)(R)))
2509
2513
 
2510
2514
 
2511
2515
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
@@ -2629,87 +2633,87 @@ _mm512_maskz_fnmsub_pd(__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
2629
2633
  }
2630
2634
 
2631
2635
  #define _mm512_fmadd_round_ps(A, B, C, R) \
2632
- (__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2633
- (__v16sf)(__m512)(B), \
2634
- (__v16sf)(__m512)(C), \
2635
- (__mmask16)-1, (int)(R))
2636
+ ((__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2637
+ (__v16sf)(__m512)(B), \
2638
+ (__v16sf)(__m512)(C), \
2639
+ (__mmask16)-1, (int)(R)))
2636
2640
 
2637
2641
 
2638
2642
  #define _mm512_mask_fmadd_round_ps(A, U, B, C, R) \
2639
- (__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2640
- (__v16sf)(__m512)(B), \
2641
- (__v16sf)(__m512)(C), \
2642
- (__mmask16)(U), (int)(R))
2643
+ ((__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2644
+ (__v16sf)(__m512)(B), \
2645
+ (__v16sf)(__m512)(C), \
2646
+ (__mmask16)(U), (int)(R)))
2643
2647
 
2644
2648
 
2645
2649
  #define _mm512_mask3_fmadd_round_ps(A, B, C, U, R) \
2646
- (__m512)__builtin_ia32_vfmaddps512_mask3((__v16sf)(__m512)(A), \
2647
- (__v16sf)(__m512)(B), \
2648
- (__v16sf)(__m512)(C), \
2649
- (__mmask16)(U), (int)(R))
2650
+ ((__m512)__builtin_ia32_vfmaddps512_mask3((__v16sf)(__m512)(A), \
2651
+ (__v16sf)(__m512)(B), \
2652
+ (__v16sf)(__m512)(C), \
2653
+ (__mmask16)(U), (int)(R)))
2650
2654
 
2651
2655
 
2652
2656
  #define _mm512_maskz_fmadd_round_ps(U, A, B, C, R) \
2653
- (__m512)__builtin_ia32_vfmaddps512_maskz((__v16sf)(__m512)(A), \
2654
- (__v16sf)(__m512)(B), \
2655
- (__v16sf)(__m512)(C), \
2656
- (__mmask16)(U), (int)(R))
2657
+ ((__m512)__builtin_ia32_vfmaddps512_maskz((__v16sf)(__m512)(A), \
2658
+ (__v16sf)(__m512)(B), \
2659
+ (__v16sf)(__m512)(C), \
2660
+ (__mmask16)(U), (int)(R)))
2657
2661
 
2658
2662
 
2659
2663
  #define _mm512_fmsub_round_ps(A, B, C, R) \
2660
- (__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2661
- (__v16sf)(__m512)(B), \
2662
- -(__v16sf)(__m512)(C), \
2663
- (__mmask16)-1, (int)(R))
2664
+ ((__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2665
+ (__v16sf)(__m512)(B), \
2666
+ -(__v16sf)(__m512)(C), \
2667
+ (__mmask16)-1, (int)(R)))
2664
2668
 
2665
2669
 
2666
2670
  #define _mm512_mask_fmsub_round_ps(A, U, B, C, R) \
2667
- (__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2668
- (__v16sf)(__m512)(B), \
2669
- -(__v16sf)(__m512)(C), \
2670
- (__mmask16)(U), (int)(R))
2671
+ ((__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2672
+ (__v16sf)(__m512)(B), \
2673
+ -(__v16sf)(__m512)(C), \
2674
+ (__mmask16)(U), (int)(R)))
2671
2675
 
2672
2676
 
2673
2677
  #define _mm512_maskz_fmsub_round_ps(U, A, B, C, R) \
2674
- (__m512)__builtin_ia32_vfmaddps512_maskz((__v16sf)(__m512)(A), \
2675
- (__v16sf)(__m512)(B), \
2676
- -(__v16sf)(__m512)(C), \
2677
- (__mmask16)(U), (int)(R))
2678
+ ((__m512)__builtin_ia32_vfmaddps512_maskz((__v16sf)(__m512)(A), \
2679
+ (__v16sf)(__m512)(B), \
2680
+ -(__v16sf)(__m512)(C), \
2681
+ (__mmask16)(U), (int)(R)))
2678
2682
 
2679
2683
 
2680
2684
  #define _mm512_fnmadd_round_ps(A, B, C, R) \
2681
- (__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2682
- -(__v16sf)(__m512)(B), \
2683
- (__v16sf)(__m512)(C), \
2684
- (__mmask16)-1, (int)(R))
2685
+ ((__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2686
+ -(__v16sf)(__m512)(B), \
2687
+ (__v16sf)(__m512)(C), \
2688
+ (__mmask16)-1, (int)(R)))
2685
2689
 
2686
2690
 
2687
2691
  #define _mm512_mask3_fnmadd_round_ps(A, B, C, U, R) \
2688
- (__m512)__builtin_ia32_vfmaddps512_mask3(-(__v16sf)(__m512)(A), \
2689
- (__v16sf)(__m512)(B), \
2690
- (__v16sf)(__m512)(C), \
2691
- (__mmask16)(U), (int)(R))
2692
+ ((__m512)__builtin_ia32_vfmaddps512_mask3(-(__v16sf)(__m512)(A), \
2693
+ (__v16sf)(__m512)(B), \
2694
+ (__v16sf)(__m512)(C), \
2695
+ (__mmask16)(U), (int)(R)))
2692
2696
 
2693
2697
 
2694
2698
  #define _mm512_maskz_fnmadd_round_ps(U, A, B, C, R) \
2695
- (__m512)__builtin_ia32_vfmaddps512_maskz(-(__v16sf)(__m512)(A), \
2696
- (__v16sf)(__m512)(B), \
2697
- (__v16sf)(__m512)(C), \
2698
- (__mmask16)(U), (int)(R))
2699
+ ((__m512)__builtin_ia32_vfmaddps512_maskz(-(__v16sf)(__m512)(A), \
2700
+ (__v16sf)(__m512)(B), \
2701
+ (__v16sf)(__m512)(C), \
2702
+ (__mmask16)(U), (int)(R)))
2699
2703
 
2700
2704
 
2701
2705
  #define _mm512_fnmsub_round_ps(A, B, C, R) \
2702
- (__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2703
- -(__v16sf)(__m512)(B), \
2704
- -(__v16sf)(__m512)(C), \
2705
- (__mmask16)-1, (int)(R))
2706
+ ((__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
2707
+ -(__v16sf)(__m512)(B), \
2708
+ -(__v16sf)(__m512)(C), \
2709
+ (__mmask16)-1, (int)(R)))
2706
2710
 
2707
2711
 
2708
2712
  #define _mm512_maskz_fnmsub_round_ps(U, A, B, C, R) \
2709
- (__m512)__builtin_ia32_vfmaddps512_maskz(-(__v16sf)(__m512)(A), \
2710
- (__v16sf)(__m512)(B), \
2711
- -(__v16sf)(__m512)(C), \
2712
- (__mmask16)(U), (int)(R))
2713
+ ((__m512)__builtin_ia32_vfmaddps512_maskz(-(__v16sf)(__m512)(A), \
2714
+ (__v16sf)(__m512)(B), \
2715
+ -(__v16sf)(__m512)(C), \
2716
+ (__mmask16)(U), (int)(R)))
2713
2717
 
2714
2718
 
2715
2719
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
@@ -2833,52 +2837,52 @@ _mm512_maskz_fnmsub_ps(__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
2833
2837
  }
2834
2838
 
2835
2839
  #define _mm512_fmaddsub_round_pd(A, B, C, R) \
2836
- (__m512d)__builtin_ia32_vfmaddsubpd512_mask((__v8df)(__m512d)(A), \
2837
- (__v8df)(__m512d)(B), \
2838
- (__v8df)(__m512d)(C), \
2839
- (__mmask8)-1, (int)(R))
2840
+ ((__m512d)__builtin_ia32_vfmaddsubpd512_mask((__v8df)(__m512d)(A), \
2841
+ (__v8df)(__m512d)(B), \
2842
+ (__v8df)(__m512d)(C), \
2843
+ (__mmask8)-1, (int)(R)))
2840
2844
 
2841
2845
 
2842
2846
  #define _mm512_mask_fmaddsub_round_pd(A, U, B, C, R) \
2843
- (__m512d)__builtin_ia32_vfmaddsubpd512_mask((__v8df)(__m512d)(A), \
2844
- (__v8df)(__m512d)(B), \
2845
- (__v8df)(__m512d)(C), \
2846
- (__mmask8)(U), (int)(R))
2847
+ ((__m512d)__builtin_ia32_vfmaddsubpd512_mask((__v8df)(__m512d)(A), \
2848
+ (__v8df)(__m512d)(B), \
2849
+ (__v8df)(__m512d)(C), \
2850
+ (__mmask8)(U), (int)(R)))
2847
2851
 
2848
2852
 
2849
2853
  #define _mm512_mask3_fmaddsub_round_pd(A, B, C, U, R) \
2850
- (__m512d)__builtin_ia32_vfmaddsubpd512_mask3((__v8df)(__m512d)(A), \
2851
- (__v8df)(__m512d)(B), \
2852
- (__v8df)(__m512d)(C), \
2853
- (__mmask8)(U), (int)(R))
2854
+ ((__m512d)__builtin_ia32_vfmaddsubpd512_mask3((__v8df)(__m512d)(A), \
2855
+ (__v8df)(__m512d)(B), \
2856
+ (__v8df)(__m512d)(C), \
2857
+ (__mmask8)(U), (int)(R)))
2854
2858
 
2855
2859
 
2856
2860
  #define _mm512_maskz_fmaddsub_round_pd(U, A, B, C, R) \
2857
- (__m512d)__builtin_ia32_vfmaddsubpd512_maskz((__v8df)(__m512d)(A), \
2858
- (__v8df)(__m512d)(B), \
2859
- (__v8df)(__m512d)(C), \
2860
- (__mmask8)(U), (int)(R))
2861
+ ((__m512d)__builtin_ia32_vfmaddsubpd512_maskz((__v8df)(__m512d)(A), \
2862
+ (__v8df)(__m512d)(B), \
2863
+ (__v8df)(__m512d)(C), \
2864
+ (__mmask8)(U), (int)(R)))
2861
2865
 
2862
2866
 
2863
2867
  #define _mm512_fmsubadd_round_pd(A, B, C, R) \
2864
- (__m512d)__builtin_ia32_vfmaddsubpd512_mask((__v8df)(__m512d)(A), \
2865
- (__v8df)(__m512d)(B), \
2866
- -(__v8df)(__m512d)(C), \
2867
- (__mmask8)-1, (int)(R))
2868
+ ((__m512d)__builtin_ia32_vfmaddsubpd512_mask((__v8df)(__m512d)(A), \
2869
+ (__v8df)(__m512d)(B), \
2870
+ -(__v8df)(__m512d)(C), \
2871
+ (__mmask8)-1, (int)(R)))
2868
2872
 
2869
2873
 
2870
2874
  #define _mm512_mask_fmsubadd_round_pd(A, U, B, C, R) \
2871
- (__m512d)__builtin_ia32_vfmaddsubpd512_mask((__v8df)(__m512d)(A), \
2872
- (__v8df)(__m512d)(B), \
2873
- -(__v8df)(__m512d)(C), \
2874
- (__mmask8)(U), (int)(R))
2875
+ ((__m512d)__builtin_ia32_vfmaddsubpd512_mask((__v8df)(__m512d)(A), \
2876
+ (__v8df)(__m512d)(B), \
2877
+ -(__v8df)(__m512d)(C), \
2878
+ (__mmask8)(U), (int)(R)))
2875
2879
 
2876
2880
 
2877
2881
  #define _mm512_maskz_fmsubadd_round_pd(U, A, B, C, R) \
2878
- (__m512d)__builtin_ia32_vfmaddsubpd512_maskz((__v8df)(__m512d)(A), \
2879
- (__v8df)(__m512d)(B), \
2880
- -(__v8df)(__m512d)(C), \
2881
- (__mmask8)(U), (int)(R))
2882
+ ((__m512d)__builtin_ia32_vfmaddsubpd512_maskz((__v8df)(__m512d)(A), \
2883
+ (__v8df)(__m512d)(B), \
2884
+ -(__v8df)(__m512d)(C), \
2885
+ (__mmask8)(U), (int)(R)))
2882
2886
 
2883
2887
 
2884
2888
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
@@ -2952,52 +2956,52 @@ _mm512_maskz_fmsubadd_pd(__mmask8 __U, __m512d __A, __m512d __B, __m512d __C)
2952
2956
  }
2953
2957
 
2954
2958
  #define _mm512_fmaddsub_round_ps(A, B, C, R) \
2955
- (__m512)__builtin_ia32_vfmaddsubps512_mask((__v16sf)(__m512)(A), \
2956
- (__v16sf)(__m512)(B), \
2957
- (__v16sf)(__m512)(C), \
2958
- (__mmask16)-1, (int)(R))
2959
+ ((__m512)__builtin_ia32_vfmaddsubps512_mask((__v16sf)(__m512)(A), \
2960
+ (__v16sf)(__m512)(B), \
2961
+ (__v16sf)(__m512)(C), \
2962
+ (__mmask16)-1, (int)(R)))
2959
2963
 
2960
2964
 
2961
2965
  #define _mm512_mask_fmaddsub_round_ps(A, U, B, C, R) \
2962
- (__m512)__builtin_ia32_vfmaddsubps512_mask((__v16sf)(__m512)(A), \
2963
- (__v16sf)(__m512)(B), \
2964
- (__v16sf)(__m512)(C), \
2965
- (__mmask16)(U), (int)(R))
2966
+ ((__m512)__builtin_ia32_vfmaddsubps512_mask((__v16sf)(__m512)(A), \
2967
+ (__v16sf)(__m512)(B), \
2968
+ (__v16sf)(__m512)(C), \
2969
+ (__mmask16)(U), (int)(R)))
2966
2970
 
2967
2971
 
2968
2972
  #define _mm512_mask3_fmaddsub_round_ps(A, B, C, U, R) \
2969
- (__m512)__builtin_ia32_vfmaddsubps512_mask3((__v16sf)(__m512)(A), \
2970
- (__v16sf)(__m512)(B), \
2971
- (__v16sf)(__m512)(C), \
2972
- (__mmask16)(U), (int)(R))
2973
+ ((__m512)__builtin_ia32_vfmaddsubps512_mask3((__v16sf)(__m512)(A), \
2974
+ (__v16sf)(__m512)(B), \
2975
+ (__v16sf)(__m512)(C), \
2976
+ (__mmask16)(U), (int)(R)))
2973
2977
 
2974
2978
 
2975
2979
  #define _mm512_maskz_fmaddsub_round_ps(U, A, B, C, R) \
2976
- (__m512)__builtin_ia32_vfmaddsubps512_maskz((__v16sf)(__m512)(A), \
2977
- (__v16sf)(__m512)(B), \
2978
- (__v16sf)(__m512)(C), \
2979
- (__mmask16)(U), (int)(R))
2980
+ ((__m512)__builtin_ia32_vfmaddsubps512_maskz((__v16sf)(__m512)(A), \
2981
+ (__v16sf)(__m512)(B), \
2982
+ (__v16sf)(__m512)(C), \
2983
+ (__mmask16)(U), (int)(R)))
2980
2984
 
2981
2985
 
2982
2986
  #define _mm512_fmsubadd_round_ps(A, B, C, R) \
2983
- (__m512)__builtin_ia32_vfmaddsubps512_mask((__v16sf)(__m512)(A), \
2984
- (__v16sf)(__m512)(B), \
2985
- -(__v16sf)(__m512)(C), \
2986
- (__mmask16)-1, (int)(R))
2987
+ ((__m512)__builtin_ia32_vfmaddsubps512_mask((__v16sf)(__m512)(A), \
2988
+ (__v16sf)(__m512)(B), \
2989
+ -(__v16sf)(__m512)(C), \
2990
+ (__mmask16)-1, (int)(R)))
2987
2991
 
2988
2992
 
2989
2993
  #define _mm512_mask_fmsubadd_round_ps(A, U, B, C, R) \
2990
- (__m512)__builtin_ia32_vfmaddsubps512_mask((__v16sf)(__m512)(A), \
2991
- (__v16sf)(__m512)(B), \
2992
- -(__v16sf)(__m512)(C), \
2993
- (__mmask16)(U), (int)(R))
2994
+ ((__m512)__builtin_ia32_vfmaddsubps512_mask((__v16sf)(__m512)(A), \
2995
+ (__v16sf)(__m512)(B), \
2996
+ -(__v16sf)(__m512)(C), \
2997
+ (__mmask16)(U), (int)(R)))
2994
2998
 
2995
2999
 
2996
3000
  #define _mm512_maskz_fmsubadd_round_ps(U, A, B, C, R) \
2997
- (__m512)__builtin_ia32_vfmaddsubps512_maskz((__v16sf)(__m512)(A), \
2998
- (__v16sf)(__m512)(B), \
2999
- -(__v16sf)(__m512)(C), \
3000
- (__mmask16)(U), (int)(R))
3001
+ ((__m512)__builtin_ia32_vfmaddsubps512_maskz((__v16sf)(__m512)(A), \
3002
+ (__v16sf)(__m512)(B), \
3003
+ -(__v16sf)(__m512)(C), \
3004
+ (__mmask16)(U), (int)(R)))
3001
3005
 
3002
3006
 
3003
3007
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
@@ -3071,10 +3075,10 @@ _mm512_maskz_fmsubadd_ps(__mmask16 __U, __m512 __A, __m512 __B, __m512 __C)
3071
3075
  }
3072
3076
 
3073
3077
  #define _mm512_mask3_fmsub_round_pd(A, B, C, U, R) \
3074
- (__m512d)__builtin_ia32_vfmsubpd512_mask3((__v8df)(__m512d)(A), \
3075
- (__v8df)(__m512d)(B), \
3076
- (__v8df)(__m512d)(C), \
3077
- (__mmask8)(U), (int)(R))
3078
+ ((__m512d)__builtin_ia32_vfmsubpd512_mask3((__v8df)(__m512d)(A), \
3079
+ (__v8df)(__m512d)(B), \
3080
+ (__v8df)(__m512d)(C), \
3081
+ (__mmask8)(U), (int)(R)))
3078
3082
 
3079
3083
 
3080
3084
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
@@ -3088,10 +3092,10 @@ _mm512_mask3_fmsub_pd(__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
3088
3092
  }
3089
3093
 
3090
3094
  #define _mm512_mask3_fmsub_round_ps(A, B, C, U, R) \
3091
- (__m512)__builtin_ia32_vfmsubps512_mask3((__v16sf)(__m512)(A), \
3092
- (__v16sf)(__m512)(B), \
3093
- (__v16sf)(__m512)(C), \
3094
- (__mmask16)(U), (int)(R))
3095
+ ((__m512)__builtin_ia32_vfmsubps512_mask3((__v16sf)(__m512)(A), \
3096
+ (__v16sf)(__m512)(B), \
3097
+ (__v16sf)(__m512)(C), \
3098
+ (__mmask16)(U), (int)(R)))
3095
3099
 
3096
3100
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
3097
3101
  _mm512_mask3_fmsub_ps(__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
@@ -3104,10 +3108,10 @@ _mm512_mask3_fmsub_ps(__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
3104
3108
  }
3105
3109
 
3106
3110
  #define _mm512_mask3_fmsubadd_round_pd(A, B, C, U, R) \
3107
- (__m512d)__builtin_ia32_vfmsubaddpd512_mask3((__v8df)(__m512d)(A), \
3108
- (__v8df)(__m512d)(B), \
3109
- (__v8df)(__m512d)(C), \
3110
- (__mmask8)(U), (int)(R))
3111
+ ((__m512d)__builtin_ia32_vfmsubaddpd512_mask3((__v8df)(__m512d)(A), \
3112
+ (__v8df)(__m512d)(B), \
3113
+ (__v8df)(__m512d)(C), \
3114
+ (__mmask8)(U), (int)(R)))
3111
3115
 
3112
3116
 
3113
3117
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
@@ -3121,10 +3125,10 @@ _mm512_mask3_fmsubadd_pd(__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
3121
3125
  }
3122
3126
 
3123
3127
  #define _mm512_mask3_fmsubadd_round_ps(A, B, C, U, R) \
3124
- (__m512)__builtin_ia32_vfmsubaddps512_mask3((__v16sf)(__m512)(A), \
3125
- (__v16sf)(__m512)(B), \
3126
- (__v16sf)(__m512)(C), \
3127
- (__mmask16)(U), (int)(R))
3128
+ ((__m512)__builtin_ia32_vfmsubaddps512_mask3((__v16sf)(__m512)(A), \
3129
+ (__v16sf)(__m512)(B), \
3130
+ (__v16sf)(__m512)(C), \
3131
+ (__mmask16)(U), (int)(R)))
3128
3132
 
3129
3133
 
3130
3134
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
@@ -3138,10 +3142,10 @@ _mm512_mask3_fmsubadd_ps(__m512 __A, __m512 __B, __m512 __C, __mmask16 __U)
3138
3142
  }
3139
3143
 
3140
3144
  #define _mm512_mask_fnmadd_round_pd(A, U, B, C, R) \
3141
- (__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
3142
- -(__v8df)(__m512d)(B), \
3143
- (__v8df)(__m512d)(C), \
3144
- (__mmask8)(U), (int)(R))
3145
+ ((__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
3146
+ -(__v8df)(__m512d)(B), \
3147
+ (__v8df)(__m512d)(C), \
3148
+ (__mmask8)(U), (int)(R)))
3145
3149
 
3146
3150
 
3147
3151
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
@@ -3155,10 +3159,10 @@ _mm512_mask_fnmadd_pd(__m512d __A, __mmask8 __U, __m512d __B, __m512d __C)
3155
3159
  }
3156
3160
 
3157
3161
  #define _mm512_mask_fnmadd_round_ps(A, U, B, C, R) \
3158
- (__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
3159
- -(__v16sf)(__m512)(B), \
3160
- (__v16sf)(__m512)(C), \
3161
- (__mmask16)(U), (int)(R))
3162
+ ((__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
3163
+ -(__v16sf)(__m512)(B), \
3164
+ (__v16sf)(__m512)(C), \
3165
+ (__mmask16)(U), (int)(R)))
3162
3166
 
3163
3167
 
3164
3168
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
@@ -3172,17 +3176,17 @@ _mm512_mask_fnmadd_ps(__m512 __A, __mmask16 __U, __m512 __B, __m512 __C)
3172
3176
  }
3173
3177
 
3174
3178
  #define _mm512_mask_fnmsub_round_pd(A, U, B, C, R) \
3175
- (__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
3176
- -(__v8df)(__m512d)(B), \
3177
- -(__v8df)(__m512d)(C), \
3178
- (__mmask8)(U), (int)(R))
3179
+ ((__m512d)__builtin_ia32_vfmaddpd512_mask((__v8df)(__m512d)(A), \
3180
+ -(__v8df)(__m512d)(B), \
3181
+ -(__v8df)(__m512d)(C), \
3182
+ (__mmask8)(U), (int)(R)))
3179
3183
 
3180
3184
 
3181
3185
  #define _mm512_mask3_fnmsub_round_pd(A, B, C, U, R) \
3182
- (__m512d)__builtin_ia32_vfmsubpd512_mask3(-(__v8df)(__m512d)(A), \
3183
- (__v8df)(__m512d)(B), \
3184
- (__v8df)(__m512d)(C), \
3185
- (__mmask8)(U), (int)(R))
3186
+ ((__m512d)__builtin_ia32_vfmsubpd512_mask3(-(__v8df)(__m512d)(A), \
3187
+ (__v8df)(__m512d)(B), \
3188
+ (__v8df)(__m512d)(C), \
3189
+ (__mmask8)(U), (int)(R)))
3186
3190
 
3187
3191
 
3188
3192
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
@@ -3206,17 +3210,17 @@ _mm512_mask3_fnmsub_pd(__m512d __A, __m512d __B, __m512d __C, __mmask8 __U)
3206
3210
  }
3207
3211
 
3208
3212
  #define _mm512_mask_fnmsub_round_ps(A, U, B, C, R) \
3209
- (__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
3210
- -(__v16sf)(__m512)(B), \
3211
- -(__v16sf)(__m512)(C), \
3212
- (__mmask16)(U), (int)(R))
3213
+ ((__m512)__builtin_ia32_vfmaddps512_mask((__v16sf)(__m512)(A), \
3214
+ -(__v16sf)(__m512)(B), \
3215
+ -(__v16sf)(__m512)(C), \
3216
+ (__mmask16)(U), (int)(R)))
3213
3217
 
3214
3218
 
3215
3219
  #define _mm512_mask3_fnmsub_round_ps(A, B, C, U, R) \
3216
- (__m512)__builtin_ia32_vfmsubps512_mask3(-(__v16sf)(__m512)(A), \
3217
- (__v16sf)(__m512)(B), \
3218
- (__v16sf)(__m512)(C), \
3219
- (__mmask16)(U), (int)(R))
3220
+ ((__m512)__builtin_ia32_vfmsubps512_mask3(-(__v16sf)(__m512)(A), \
3221
+ (__v16sf)(__m512)(B), \
3222
+ (__v16sf)(__m512)(C), \
3223
+ (__mmask16)(U), (int)(R)))
3220
3224
 
3221
3225
 
3222
3226
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
@@ -3312,63 +3316,63 @@ _mm512_maskz_permutex2var_epi64(__mmask8 __U, __m512i __A, __m512i __I,
3312
3316
  }
3313
3317
 
3314
3318
  #define _mm512_alignr_epi64(A, B, I) \
3315
- (__m512i)__builtin_ia32_alignq512((__v8di)(__m512i)(A), \
3316
- (__v8di)(__m512i)(B), (int)(I))
3319
+ ((__m512i)__builtin_ia32_alignq512((__v8di)(__m512i)(A), \
3320
+ (__v8di)(__m512i)(B), (int)(I)))
3317
3321
 
3318
3322
  #define _mm512_mask_alignr_epi64(W, U, A, B, imm) \
3319
- (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
3320
- (__v8di)_mm512_alignr_epi64((A), (B), (imm)), \
3321
- (__v8di)(__m512i)(W))
3323
+ ((__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
3324
+ (__v8di)_mm512_alignr_epi64((A), (B), (imm)), \
3325
+ (__v8di)(__m512i)(W)))
3322
3326
 
3323
3327
  #define _mm512_maskz_alignr_epi64(U, A, B, imm) \
3324
- (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
3325
- (__v8di)_mm512_alignr_epi64((A), (B), (imm)), \
3326
- (__v8di)_mm512_setzero_si512())
3328
+ ((__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
3329
+ (__v8di)_mm512_alignr_epi64((A), (B), (imm)), \
3330
+ (__v8di)_mm512_setzero_si512()))
3327
3331
 
3328
3332
  #define _mm512_alignr_epi32(A, B, I) \
3329
- (__m512i)__builtin_ia32_alignd512((__v16si)(__m512i)(A), \
3330
- (__v16si)(__m512i)(B), (int)(I))
3333
+ ((__m512i)__builtin_ia32_alignd512((__v16si)(__m512i)(A), \
3334
+ (__v16si)(__m512i)(B), (int)(I)))
3331
3335
 
3332
3336
  #define _mm512_mask_alignr_epi32(W, U, A, B, imm) \
3333
- (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
3334
- (__v16si)_mm512_alignr_epi32((A), (B), (imm)), \
3335
- (__v16si)(__m512i)(W))
3337
+ ((__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
3338
+ (__v16si)_mm512_alignr_epi32((A), (B), (imm)), \
3339
+ (__v16si)(__m512i)(W)))
3336
3340
 
3337
3341
  #define _mm512_maskz_alignr_epi32(U, A, B, imm) \
3338
- (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
3339
- (__v16si)_mm512_alignr_epi32((A), (B), (imm)), \
3340
- (__v16si)_mm512_setzero_si512())
3342
+ ((__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
3343
+ (__v16si)_mm512_alignr_epi32((A), (B), (imm)), \
3344
+ (__v16si)_mm512_setzero_si512()))
3341
3345
  /* Vector Extract */
3342
3346
 
3343
3347
  #define _mm512_extractf64x4_pd(A, I) \
3344
- (__m256d)__builtin_ia32_extractf64x4_mask((__v8df)(__m512d)(A), (int)(I), \
3345
- (__v4df)_mm256_undefined_pd(), \
3346
- (__mmask8)-1)
3348
+ ((__m256d)__builtin_ia32_extractf64x4_mask((__v8df)(__m512d)(A), (int)(I), \
3349
+ (__v4df)_mm256_undefined_pd(), \
3350
+ (__mmask8)-1))
3347
3351
 
3348
3352
  #define _mm512_mask_extractf64x4_pd(W, U, A, imm) \
3349
- (__m256d)__builtin_ia32_extractf64x4_mask((__v8df)(__m512d)(A), (int)(imm), \
3350
- (__v4df)(__m256d)(W), \
3351
- (__mmask8)(U))
3353
+ ((__m256d)__builtin_ia32_extractf64x4_mask((__v8df)(__m512d)(A), (int)(imm), \
3354
+ (__v4df)(__m256d)(W), \
3355
+ (__mmask8)(U)))
3352
3356
 
3353
3357
  #define _mm512_maskz_extractf64x4_pd(U, A, imm) \
3354
- (__m256d)__builtin_ia32_extractf64x4_mask((__v8df)(__m512d)(A), (int)(imm), \
3355
- (__v4df)_mm256_setzero_pd(), \
3356
- (__mmask8)(U))
3358
+ ((__m256d)__builtin_ia32_extractf64x4_mask((__v8df)(__m512d)(A), (int)(imm), \
3359
+ (__v4df)_mm256_setzero_pd(), \
3360
+ (__mmask8)(U)))
3357
3361
 
3358
3362
  #define _mm512_extractf32x4_ps(A, I) \
3359
- (__m128)__builtin_ia32_extractf32x4_mask((__v16sf)(__m512)(A), (int)(I), \
3360
- (__v4sf)_mm_undefined_ps(), \
3361
- (__mmask8)-1)
3363
+ ((__m128)__builtin_ia32_extractf32x4_mask((__v16sf)(__m512)(A), (int)(I), \
3364
+ (__v4sf)_mm_undefined_ps(), \
3365
+ (__mmask8)-1))
3362
3366
 
3363
3367
  #define _mm512_mask_extractf32x4_ps(W, U, A, imm) \
3364
- (__m128)__builtin_ia32_extractf32x4_mask((__v16sf)(__m512)(A), (int)(imm), \
3365
- (__v4sf)(__m128)(W), \
3366
- (__mmask8)(U))
3368
+ ((__m128)__builtin_ia32_extractf32x4_mask((__v16sf)(__m512)(A), (int)(imm), \
3369
+ (__v4sf)(__m128)(W), \
3370
+ (__mmask8)(U)))
3367
3371
 
3368
3372
  #define _mm512_maskz_extractf32x4_ps(U, A, imm) \
3369
- (__m128)__builtin_ia32_extractf32x4_mask((__v16sf)(__m512)(A), (int)(imm), \
3370
- (__v4sf)_mm_setzero_ps(), \
3371
- (__mmask8)(U))
3373
+ ((__m128)__builtin_ia32_extractf32x4_mask((__v16sf)(__m512)(A), (int)(imm), \
3374
+ (__v4sf)_mm_setzero_ps(), \
3375
+ (__mmask8)(U)))
3372
3376
 
3373
3377
  /* Vector Blend */
3374
3378
 
@@ -3407,14 +3411,14 @@ _mm512_mask_blend_epi32(__mmask16 __U, __m512i __A, __m512i __W)
3407
3411
  /* Compare */
3408
3412
 
3409
3413
  #define _mm512_cmp_round_ps_mask(A, B, P, R) \
3410
- (__mmask16)__builtin_ia32_cmpps512_mask((__v16sf)(__m512)(A), \
3411
- (__v16sf)(__m512)(B), (int)(P), \
3412
- (__mmask16)-1, (int)(R))
3414
+ ((__mmask16)__builtin_ia32_cmpps512_mask((__v16sf)(__m512)(A), \
3415
+ (__v16sf)(__m512)(B), (int)(P), \
3416
+ (__mmask16)-1, (int)(R)))
3413
3417
 
3414
3418
  #define _mm512_mask_cmp_round_ps_mask(U, A, B, P, R) \
3415
- (__mmask16)__builtin_ia32_cmpps512_mask((__v16sf)(__m512)(A), \
3416
- (__v16sf)(__m512)(B), (int)(P), \
3417
- (__mmask16)(U), (int)(R))
3419
+ ((__mmask16)__builtin_ia32_cmpps512_mask((__v16sf)(__m512)(A), \
3420
+ (__v16sf)(__m512)(B), (int)(P), \
3421
+ (__mmask16)(U), (int)(R)))
3418
3422
 
3419
3423
  #define _mm512_cmp_ps_mask(A, B, P) \
3420
3424
  _mm512_cmp_round_ps_mask((A), (B), (P), _MM_FROUND_CUR_DIRECTION)
@@ -3462,14 +3466,14 @@ _mm512_mask_blend_epi32(__mmask16 __U, __m512i __A, __m512i __W)
3462
3466
  _mm512_mask_cmp_ps_mask((k), (A), (B), _CMP_ORD_Q)
3463
3467
 
3464
3468
  #define _mm512_cmp_round_pd_mask(A, B, P, R) \
3465
- (__mmask8)__builtin_ia32_cmppd512_mask((__v8df)(__m512d)(A), \
3466
- (__v8df)(__m512d)(B), (int)(P), \
3467
- (__mmask8)-1, (int)(R))
3469
+ ((__mmask8)__builtin_ia32_cmppd512_mask((__v8df)(__m512d)(A), \
3470
+ (__v8df)(__m512d)(B), (int)(P), \
3471
+ (__mmask8)-1, (int)(R)))
3468
3472
 
3469
3473
  #define _mm512_mask_cmp_round_pd_mask(U, A, B, P, R) \
3470
- (__mmask8)__builtin_ia32_cmppd512_mask((__v8df)(__m512d)(A), \
3471
- (__v8df)(__m512d)(B), (int)(P), \
3472
- (__mmask8)(U), (int)(R))
3474
+ ((__mmask8)__builtin_ia32_cmppd512_mask((__v8df)(__m512d)(A), \
3475
+ (__v8df)(__m512d)(B), (int)(P), \
3476
+ (__mmask8)(U), (int)(R)))
3473
3477
 
3474
3478
  #define _mm512_cmp_pd_mask(A, B, P) \
3475
3479
  _mm512_cmp_round_pd_mask((A), (B), (P), _MM_FROUND_CUR_DIRECTION)
@@ -3519,19 +3523,19 @@ _mm512_mask_blend_epi32(__mmask16 __U, __m512i __A, __m512i __W)
3519
3523
  /* Conversion */
3520
3524
 
3521
3525
  #define _mm512_cvtt_roundps_epu32(A, R) \
3522
- (__m512i)__builtin_ia32_cvttps2udq512_mask((__v16sf)(__m512)(A), \
3523
- (__v16si)_mm512_undefined_epi32(), \
3524
- (__mmask16)-1, (int)(R))
3526
+ ((__m512i)__builtin_ia32_cvttps2udq512_mask((__v16sf)(__m512)(A), \
3527
+ (__v16si)_mm512_undefined_epi32(), \
3528
+ (__mmask16)-1, (int)(R)))
3525
3529
 
3526
3530
  #define _mm512_mask_cvtt_roundps_epu32(W, U, A, R) \
3527
- (__m512i)__builtin_ia32_cvttps2udq512_mask((__v16sf)(__m512)(A), \
3528
- (__v16si)(__m512i)(W), \
3529
- (__mmask16)(U), (int)(R))
3531
+ ((__m512i)__builtin_ia32_cvttps2udq512_mask((__v16sf)(__m512)(A), \
3532
+ (__v16si)(__m512i)(W), \
3533
+ (__mmask16)(U), (int)(R)))
3530
3534
 
3531
3535
  #define _mm512_maskz_cvtt_roundps_epu32(U, A, R) \
3532
- (__m512i)__builtin_ia32_cvttps2udq512_mask((__v16sf)(__m512)(A), \
3533
- (__v16si)_mm512_setzero_si512(), \
3534
- (__mmask16)(U), (int)(R))
3536
+ ((__m512i)__builtin_ia32_cvttps2udq512_mask((__v16sf)(__m512)(A), \
3537
+ (__v16si)_mm512_setzero_si512(), \
3538
+ (__mmask16)(U), (int)(R)))
3535
3539
 
3536
3540
 
3537
3541
  static __inline __m512i __DEFAULT_FN_ATTRS512
@@ -3563,34 +3567,34 @@ _mm512_maskz_cvttps_epu32 (__mmask16 __U, __m512 __A)
3563
3567
  }
3564
3568
 
3565
3569
  #define _mm512_cvt_roundepi32_ps(A, R) \
3566
- (__m512)__builtin_ia32_cvtdq2ps512_mask((__v16si)(__m512i)(A), \
3567
- (__v16sf)_mm512_setzero_ps(), \
3568
- (__mmask16)-1, (int)(R))
3570
+ ((__m512)__builtin_ia32_cvtdq2ps512_mask((__v16si)(__m512i)(A), \
3571
+ (__v16sf)_mm512_setzero_ps(), \
3572
+ (__mmask16)-1, (int)(R)))
3569
3573
 
3570
3574
  #define _mm512_mask_cvt_roundepi32_ps(W, U, A, R) \
3571
- (__m512)__builtin_ia32_cvtdq2ps512_mask((__v16si)(__m512i)(A), \
3572
- (__v16sf)(__m512)(W), \
3573
- (__mmask16)(U), (int)(R))
3575
+ ((__m512)__builtin_ia32_cvtdq2ps512_mask((__v16si)(__m512i)(A), \
3576
+ (__v16sf)(__m512)(W), \
3577
+ (__mmask16)(U), (int)(R)))
3574
3578
 
3575
3579
  #define _mm512_maskz_cvt_roundepi32_ps(U, A, R) \
3576
- (__m512)__builtin_ia32_cvtdq2ps512_mask((__v16si)(__m512i)(A), \
3577
- (__v16sf)_mm512_setzero_ps(), \
3578
- (__mmask16)(U), (int)(R))
3580
+ ((__m512)__builtin_ia32_cvtdq2ps512_mask((__v16si)(__m512i)(A), \
3581
+ (__v16sf)_mm512_setzero_ps(), \
3582
+ (__mmask16)(U), (int)(R)))
3579
3583
 
3580
3584
  #define _mm512_cvt_roundepu32_ps(A, R) \
3581
- (__m512)__builtin_ia32_cvtudq2ps512_mask((__v16si)(__m512i)(A), \
3582
- (__v16sf)_mm512_setzero_ps(), \
3583
- (__mmask16)-1, (int)(R))
3585
+ ((__m512)__builtin_ia32_cvtudq2ps512_mask((__v16si)(__m512i)(A), \
3586
+ (__v16sf)_mm512_setzero_ps(), \
3587
+ (__mmask16)-1, (int)(R)))
3584
3588
 
3585
3589
  #define _mm512_mask_cvt_roundepu32_ps(W, U, A, R) \
3586
- (__m512)__builtin_ia32_cvtudq2ps512_mask((__v16si)(__m512i)(A), \
3587
- (__v16sf)(__m512)(W), \
3588
- (__mmask16)(U), (int)(R))
3590
+ ((__m512)__builtin_ia32_cvtudq2ps512_mask((__v16si)(__m512i)(A), \
3591
+ (__v16sf)(__m512)(W), \
3592
+ (__mmask16)(U), (int)(R)))
3589
3593
 
3590
3594
  #define _mm512_maskz_cvt_roundepu32_ps(U, A, R) \
3591
- (__m512)__builtin_ia32_cvtudq2ps512_mask((__v16si)(__m512i)(A), \
3592
- (__v16sf)_mm512_setzero_ps(), \
3593
- (__mmask16)(U), (int)(R))
3595
+ ((__m512)__builtin_ia32_cvtudq2ps512_mask((__v16si)(__m512i)(A), \
3596
+ (__v16sf)_mm512_setzero_ps(), \
3597
+ (__mmask16)(U), (int)(R)))
3594
3598
 
3595
3599
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
3596
3600
  _mm512_cvtepu32_ps (__m512i __A)
@@ -3705,19 +3709,19 @@ _mm512_mask_cvtepu32lo_pd(__m512d __W, __mmask8 __U,__m512i __A)
3705
3709
  }
3706
3710
 
3707
3711
  #define _mm512_cvt_roundpd_ps(A, R) \
3708
- (__m256)__builtin_ia32_cvtpd2ps512_mask((__v8df)(__m512d)(A), \
3709
- (__v8sf)_mm256_setzero_ps(), \
3710
- (__mmask8)-1, (int)(R))
3712
+ ((__m256)__builtin_ia32_cvtpd2ps512_mask((__v8df)(__m512d)(A), \
3713
+ (__v8sf)_mm256_setzero_ps(), \
3714
+ (__mmask8)-1, (int)(R)))
3711
3715
 
3712
3716
  #define _mm512_mask_cvt_roundpd_ps(W, U, A, R) \
3713
- (__m256)__builtin_ia32_cvtpd2ps512_mask((__v8df)(__m512d)(A), \
3714
- (__v8sf)(__m256)(W), (__mmask8)(U), \
3715
- (int)(R))
3717
+ ((__m256)__builtin_ia32_cvtpd2ps512_mask((__v8df)(__m512d)(A), \
3718
+ (__v8sf)(__m256)(W), (__mmask8)(U), \
3719
+ (int)(R)))
3716
3720
 
3717
3721
  #define _mm512_maskz_cvt_roundpd_ps(U, A, R) \
3718
- (__m256)__builtin_ia32_cvtpd2ps512_mask((__v8df)(__m512d)(A), \
3719
- (__v8sf)_mm256_setzero_ps(), \
3720
- (__mmask8)(U), (int)(R))
3722
+ ((__m256)__builtin_ia32_cvtpd2ps512_mask((__v8df)(__m512d)(A), \
3723
+ (__v8sf)_mm256_setzero_ps(), \
3724
+ (__mmask8)(U), (int)(R)))
3721
3725
 
3722
3726
  static __inline__ __m256 __DEFAULT_FN_ATTRS512
3723
3727
  _mm512_cvtpd_ps (__m512d __A)
@@ -3765,38 +3769,38 @@ _mm512_mask_cvtpd_pslo (__m512 __W, __mmask8 __U,__m512d __A)
3765
3769
  }
3766
3770
 
3767
3771
  #define _mm512_cvt_roundps_ph(A, I) \
3768
- (__m256i)__builtin_ia32_vcvtps2ph512_mask((__v16sf)(__m512)(A), (int)(I), \
3769
- (__v16hi)_mm256_undefined_si256(), \
3770
- (__mmask16)-1)
3772
+ ((__m256i)__builtin_ia32_vcvtps2ph512_mask((__v16sf)(__m512)(A), (int)(I), \
3773
+ (__v16hi)_mm256_undefined_si256(), \
3774
+ (__mmask16)-1))
3771
3775
 
3772
3776
  #define _mm512_mask_cvt_roundps_ph(U, W, A, I) \
3773
- (__m256i)__builtin_ia32_vcvtps2ph512_mask((__v16sf)(__m512)(A), (int)(I), \
3774
- (__v16hi)(__m256i)(U), \
3775
- (__mmask16)(W))
3777
+ ((__m256i)__builtin_ia32_vcvtps2ph512_mask((__v16sf)(__m512)(A), (int)(I), \
3778
+ (__v16hi)(__m256i)(U), \
3779
+ (__mmask16)(W)))
3776
3780
 
3777
3781
  #define _mm512_maskz_cvt_roundps_ph(W, A, I) \
3778
- (__m256i)__builtin_ia32_vcvtps2ph512_mask((__v16sf)(__m512)(A), (int)(I), \
3779
- (__v16hi)_mm256_setzero_si256(), \
3780
- (__mmask16)(W))
3782
+ ((__m256i)__builtin_ia32_vcvtps2ph512_mask((__v16sf)(__m512)(A), (int)(I), \
3783
+ (__v16hi)_mm256_setzero_si256(), \
3784
+ (__mmask16)(W)))
3781
3785
 
3782
3786
  #define _mm512_cvtps_ph _mm512_cvt_roundps_ph
3783
3787
  #define _mm512_mask_cvtps_ph _mm512_mask_cvt_roundps_ph
3784
3788
  #define _mm512_maskz_cvtps_ph _mm512_maskz_cvt_roundps_ph
3785
3789
 
3786
3790
  #define _mm512_cvt_roundph_ps(A, R) \
3787
- (__m512)__builtin_ia32_vcvtph2ps512_mask((__v16hi)(__m256i)(A), \
3788
- (__v16sf)_mm512_undefined_ps(), \
3789
- (__mmask16)-1, (int)(R))
3791
+ ((__m512)__builtin_ia32_vcvtph2ps512_mask((__v16hi)(__m256i)(A), \
3792
+ (__v16sf)_mm512_undefined_ps(), \
3793
+ (__mmask16)-1, (int)(R)))
3790
3794
 
3791
3795
  #define _mm512_mask_cvt_roundph_ps(W, U, A, R) \
3792
- (__m512)__builtin_ia32_vcvtph2ps512_mask((__v16hi)(__m256i)(A), \
3793
- (__v16sf)(__m512)(W), \
3794
- (__mmask16)(U), (int)(R))
3796
+ ((__m512)__builtin_ia32_vcvtph2ps512_mask((__v16hi)(__m256i)(A), \
3797
+ (__v16sf)(__m512)(W), \
3798
+ (__mmask16)(U), (int)(R)))
3795
3799
 
3796
3800
  #define _mm512_maskz_cvt_roundph_ps(U, A, R) \
3797
- (__m512)__builtin_ia32_vcvtph2ps512_mask((__v16hi)(__m256i)(A), \
3798
- (__v16sf)_mm512_setzero_ps(), \
3799
- (__mmask16)(U), (int)(R))
3801
+ ((__m512)__builtin_ia32_vcvtph2ps512_mask((__v16hi)(__m256i)(A), \
3802
+ (__v16sf)_mm512_setzero_ps(), \
3803
+ (__mmask16)(U), (int)(R)))
3800
3804
 
3801
3805
 
3802
3806
  static __inline __m512 __DEFAULT_FN_ATTRS512
@@ -3828,19 +3832,19 @@ _mm512_maskz_cvtph_ps (__mmask16 __U, __m256i __A)
3828
3832
  }
3829
3833
 
3830
3834
  #define _mm512_cvtt_roundpd_epi32(A, R) \
3831
- (__m256i)__builtin_ia32_cvttpd2dq512_mask((__v8df)(__m512d)(A), \
3832
- (__v8si)_mm256_setzero_si256(), \
3833
- (__mmask8)-1, (int)(R))
3835
+ ((__m256i)__builtin_ia32_cvttpd2dq512_mask((__v8df)(__m512d)(A), \
3836
+ (__v8si)_mm256_setzero_si256(), \
3837
+ (__mmask8)-1, (int)(R)))
3834
3838
 
3835
3839
  #define _mm512_mask_cvtt_roundpd_epi32(W, U, A, R) \
3836
- (__m256i)__builtin_ia32_cvttpd2dq512_mask((__v8df)(__m512d)(A), \
3837
- (__v8si)(__m256i)(W), \
3838
- (__mmask8)(U), (int)(R))
3840
+ ((__m256i)__builtin_ia32_cvttpd2dq512_mask((__v8df)(__m512d)(A), \
3841
+ (__v8si)(__m256i)(W), \
3842
+ (__mmask8)(U), (int)(R)))
3839
3843
 
3840
3844
  #define _mm512_maskz_cvtt_roundpd_epi32(U, A, R) \
3841
- (__m256i)__builtin_ia32_cvttpd2dq512_mask((__v8df)(__m512d)(A), \
3842
- (__v8si)_mm256_setzero_si256(), \
3843
- (__mmask8)(U), (int)(R))
3845
+ ((__m256i)__builtin_ia32_cvttpd2dq512_mask((__v8df)(__m512d)(A), \
3846
+ (__v8si)_mm256_setzero_si256(), \
3847
+ (__mmask8)(U), (int)(R)))
3844
3848
 
3845
3849
  static __inline __m256i __DEFAULT_FN_ATTRS512
3846
3850
  _mm512_cvttpd_epi32(__m512d __a)
@@ -3870,19 +3874,19 @@ _mm512_maskz_cvttpd_epi32 (__mmask8 __U, __m512d __A)
3870
3874
  }
3871
3875
 
3872
3876
  #define _mm512_cvtt_roundps_epi32(A, R) \
3873
- (__m512i)__builtin_ia32_cvttps2dq512_mask((__v16sf)(__m512)(A), \
3874
- (__v16si)_mm512_setzero_si512(), \
3875
- (__mmask16)-1, (int)(R))
3877
+ ((__m512i)__builtin_ia32_cvttps2dq512_mask((__v16sf)(__m512)(A), \
3878
+ (__v16si)_mm512_setzero_si512(), \
3879
+ (__mmask16)-1, (int)(R)))
3876
3880
 
3877
3881
  #define _mm512_mask_cvtt_roundps_epi32(W, U, A, R) \
3878
- (__m512i)__builtin_ia32_cvttps2dq512_mask((__v16sf)(__m512)(A), \
3879
- (__v16si)(__m512i)(W), \
3880
- (__mmask16)(U), (int)(R))
3882
+ ((__m512i)__builtin_ia32_cvttps2dq512_mask((__v16sf)(__m512)(A), \
3883
+ (__v16si)(__m512i)(W), \
3884
+ (__mmask16)(U), (int)(R)))
3881
3885
 
3882
3886
  #define _mm512_maskz_cvtt_roundps_epi32(U, A, R) \
3883
- (__m512i)__builtin_ia32_cvttps2dq512_mask((__v16sf)(__m512)(A), \
3884
- (__v16si)_mm512_setzero_si512(), \
3885
- (__mmask16)(U), (int)(R))
3887
+ ((__m512i)__builtin_ia32_cvttps2dq512_mask((__v16sf)(__m512)(A), \
3888
+ (__v16si)_mm512_setzero_si512(), \
3889
+ (__mmask16)(U), (int)(R)))
3886
3890
 
3887
3891
  static __inline __m512i __DEFAULT_FN_ATTRS512
3888
3892
  _mm512_cvttps_epi32(__m512 __a)
@@ -3912,19 +3916,19 @@ _mm512_maskz_cvttps_epi32 (__mmask16 __U, __m512 __A)
3912
3916
  }
3913
3917
 
3914
3918
  #define _mm512_cvt_roundps_epi32(A, R) \
3915
- (__m512i)__builtin_ia32_cvtps2dq512_mask((__v16sf)(__m512)(A), \
3916
- (__v16si)_mm512_setzero_si512(), \
3917
- (__mmask16)-1, (int)(R))
3919
+ ((__m512i)__builtin_ia32_cvtps2dq512_mask((__v16sf)(__m512)(A), \
3920
+ (__v16si)_mm512_setzero_si512(), \
3921
+ (__mmask16)-1, (int)(R)))
3918
3922
 
3919
3923
  #define _mm512_mask_cvt_roundps_epi32(W, U, A, R) \
3920
- (__m512i)__builtin_ia32_cvtps2dq512_mask((__v16sf)(__m512)(A), \
3921
- (__v16si)(__m512i)(W), \
3922
- (__mmask16)(U), (int)(R))
3924
+ ((__m512i)__builtin_ia32_cvtps2dq512_mask((__v16sf)(__m512)(A), \
3925
+ (__v16si)(__m512i)(W), \
3926
+ (__mmask16)(U), (int)(R)))
3923
3927
 
3924
3928
  #define _mm512_maskz_cvt_roundps_epi32(U, A, R) \
3925
- (__m512i)__builtin_ia32_cvtps2dq512_mask((__v16sf)(__m512)(A), \
3926
- (__v16si)_mm512_setzero_si512(), \
3927
- (__mmask16)(U), (int)(R))
3929
+ ((__m512i)__builtin_ia32_cvtps2dq512_mask((__v16sf)(__m512)(A), \
3930
+ (__v16si)_mm512_setzero_si512(), \
3931
+ (__mmask16)(U), (int)(R)))
3928
3932
 
3929
3933
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
3930
3934
  _mm512_cvtps_epi32 (__m512 __A)
@@ -3955,19 +3959,19 @@ _mm512_maskz_cvtps_epi32 (__mmask16 __U, __m512 __A)
3955
3959
  }
3956
3960
 
3957
3961
  #define _mm512_cvt_roundpd_epi32(A, R) \
3958
- (__m256i)__builtin_ia32_cvtpd2dq512_mask((__v8df)(__m512d)(A), \
3959
- (__v8si)_mm256_setzero_si256(), \
3960
- (__mmask8)-1, (int)(R))
3962
+ ((__m256i)__builtin_ia32_cvtpd2dq512_mask((__v8df)(__m512d)(A), \
3963
+ (__v8si)_mm256_setzero_si256(), \
3964
+ (__mmask8)-1, (int)(R)))
3961
3965
 
3962
3966
  #define _mm512_mask_cvt_roundpd_epi32(W, U, A, R) \
3963
- (__m256i)__builtin_ia32_cvtpd2dq512_mask((__v8df)(__m512d)(A), \
3964
- (__v8si)(__m256i)(W), \
3965
- (__mmask8)(U), (int)(R))
3967
+ ((__m256i)__builtin_ia32_cvtpd2dq512_mask((__v8df)(__m512d)(A), \
3968
+ (__v8si)(__m256i)(W), \
3969
+ (__mmask8)(U), (int)(R)))
3966
3970
 
3967
3971
  #define _mm512_maskz_cvt_roundpd_epi32(U, A, R) \
3968
- (__m256i)__builtin_ia32_cvtpd2dq512_mask((__v8df)(__m512d)(A), \
3969
- (__v8si)_mm256_setzero_si256(), \
3970
- (__mmask8)(U), (int)(R))
3972
+ ((__m256i)__builtin_ia32_cvtpd2dq512_mask((__v8df)(__m512d)(A), \
3973
+ (__v8si)_mm256_setzero_si256(), \
3974
+ (__mmask8)(U), (int)(R)))
3971
3975
 
3972
3976
  static __inline__ __m256i __DEFAULT_FN_ATTRS512
3973
3977
  _mm512_cvtpd_epi32 (__m512d __A)
@@ -3999,19 +4003,19 @@ _mm512_maskz_cvtpd_epi32 (__mmask8 __U, __m512d __A)
3999
4003
  }
4000
4004
 
4001
4005
  #define _mm512_cvt_roundps_epu32(A, R) \
4002
- (__m512i)__builtin_ia32_cvtps2udq512_mask((__v16sf)(__m512)(A), \
4003
- (__v16si)_mm512_setzero_si512(), \
4004
- (__mmask16)-1, (int)(R))
4006
+ ((__m512i)__builtin_ia32_cvtps2udq512_mask((__v16sf)(__m512)(A), \
4007
+ (__v16si)_mm512_setzero_si512(), \
4008
+ (__mmask16)-1, (int)(R)))
4005
4009
 
4006
4010
  #define _mm512_mask_cvt_roundps_epu32(W, U, A, R) \
4007
- (__m512i)__builtin_ia32_cvtps2udq512_mask((__v16sf)(__m512)(A), \
4008
- (__v16si)(__m512i)(W), \
4009
- (__mmask16)(U), (int)(R))
4011
+ ((__m512i)__builtin_ia32_cvtps2udq512_mask((__v16sf)(__m512)(A), \
4012
+ (__v16si)(__m512i)(W), \
4013
+ (__mmask16)(U), (int)(R)))
4010
4014
 
4011
4015
  #define _mm512_maskz_cvt_roundps_epu32(U, A, R) \
4012
- (__m512i)__builtin_ia32_cvtps2udq512_mask((__v16sf)(__m512)(A), \
4013
- (__v16si)_mm512_setzero_si512(), \
4014
- (__mmask16)(U), (int)(R))
4016
+ ((__m512i)__builtin_ia32_cvtps2udq512_mask((__v16sf)(__m512)(A), \
4017
+ (__v16si)_mm512_setzero_si512(), \
4018
+ (__mmask16)(U), (int)(R)))
4015
4019
 
4016
4020
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
4017
4021
  _mm512_cvtps_epu32 ( __m512 __A)
@@ -4043,19 +4047,19 @@ _mm512_maskz_cvtps_epu32 ( __mmask16 __U, __m512 __A)
4043
4047
  }
4044
4048
 
4045
4049
  #define _mm512_cvt_roundpd_epu32(A, R) \
4046
- (__m256i)__builtin_ia32_cvtpd2udq512_mask((__v8df)(__m512d)(A), \
4047
- (__v8si)_mm256_setzero_si256(), \
4048
- (__mmask8)-1, (int)(R))
4050
+ ((__m256i)__builtin_ia32_cvtpd2udq512_mask((__v8df)(__m512d)(A), \
4051
+ (__v8si)_mm256_setzero_si256(), \
4052
+ (__mmask8)-1, (int)(R)))
4049
4053
 
4050
4054
  #define _mm512_mask_cvt_roundpd_epu32(W, U, A, R) \
4051
- (__m256i)__builtin_ia32_cvtpd2udq512_mask((__v8df)(__m512d)(A), \
4052
- (__v8si)(__m256i)(W), \
4053
- (__mmask8)(U), (int)(R))
4055
+ ((__m256i)__builtin_ia32_cvtpd2udq512_mask((__v8df)(__m512d)(A), \
4056
+ (__v8si)(__m256i)(W), \
4057
+ (__mmask8)(U), (int)(R)))
4054
4058
 
4055
4059
  #define _mm512_maskz_cvt_roundpd_epu32(U, A, R) \
4056
- (__m256i)__builtin_ia32_cvtpd2udq512_mask((__v8df)(__m512d)(A), \
4057
- (__v8si)_mm256_setzero_si256(), \
4058
- (__mmask8)(U), (int)(R))
4060
+ ((__m256i)__builtin_ia32_cvtpd2udq512_mask((__v8df)(__m512d)(A), \
4061
+ (__v8si)_mm256_setzero_si256(), \
4062
+ (__mmask8)(U), (int)(R)))
4059
4063
 
4060
4064
  static __inline__ __m256i __DEFAULT_FN_ATTRS512
4061
4065
  _mm512_cvtpd_epu32 (__m512d __A)
@@ -4975,70 +4979,70 @@ _mm512_maskz_rorv_epi64 (__mmask8 __U, __m512i __A, __m512i __B)
4975
4979
 
4976
4980
 
4977
4981
  #define _mm512_cmp_epi32_mask(a, b, p) \
4978
- (__mmask16)__builtin_ia32_cmpd512_mask((__v16si)(__m512i)(a), \
4979
- (__v16si)(__m512i)(b), (int)(p), \
4980
- (__mmask16)-1)
4982
+ ((__mmask16)__builtin_ia32_cmpd512_mask((__v16si)(__m512i)(a), \
4983
+ (__v16si)(__m512i)(b), (int)(p), \
4984
+ (__mmask16)-1))
4981
4985
 
4982
4986
  #define _mm512_cmp_epu32_mask(a, b, p) \
4983
- (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)(__m512i)(a), \
4984
- (__v16si)(__m512i)(b), (int)(p), \
4985
- (__mmask16)-1)
4987
+ ((__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)(__m512i)(a), \
4988
+ (__v16si)(__m512i)(b), (int)(p), \
4989
+ (__mmask16)-1))
4986
4990
 
4987
4991
  #define _mm512_cmp_epi64_mask(a, b, p) \
4988
- (__mmask8)__builtin_ia32_cmpq512_mask((__v8di)(__m512i)(a), \
4989
- (__v8di)(__m512i)(b), (int)(p), \
4990
- (__mmask8)-1)
4992
+ ((__mmask8)__builtin_ia32_cmpq512_mask((__v8di)(__m512i)(a), \
4993
+ (__v8di)(__m512i)(b), (int)(p), \
4994
+ (__mmask8)-1))
4991
4995
 
4992
4996
  #define _mm512_cmp_epu64_mask(a, b, p) \
4993
- (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)(__m512i)(a), \
4994
- (__v8di)(__m512i)(b), (int)(p), \
4995
- (__mmask8)-1)
4997
+ ((__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)(__m512i)(a), \
4998
+ (__v8di)(__m512i)(b), (int)(p), \
4999
+ (__mmask8)-1))
4996
5000
 
4997
5001
  #define _mm512_mask_cmp_epi32_mask(m, a, b, p) \
4998
- (__mmask16)__builtin_ia32_cmpd512_mask((__v16si)(__m512i)(a), \
4999
- (__v16si)(__m512i)(b), (int)(p), \
5000
- (__mmask16)(m))
5002
+ ((__mmask16)__builtin_ia32_cmpd512_mask((__v16si)(__m512i)(a), \
5003
+ (__v16si)(__m512i)(b), (int)(p), \
5004
+ (__mmask16)(m)))
5001
5005
 
5002
5006
  #define _mm512_mask_cmp_epu32_mask(m, a, b, p) \
5003
- (__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)(__m512i)(a), \
5004
- (__v16si)(__m512i)(b), (int)(p), \
5005
- (__mmask16)(m))
5007
+ ((__mmask16)__builtin_ia32_ucmpd512_mask((__v16si)(__m512i)(a), \
5008
+ (__v16si)(__m512i)(b), (int)(p), \
5009
+ (__mmask16)(m)))
5006
5010
 
5007
5011
  #define _mm512_mask_cmp_epi64_mask(m, a, b, p) \
5008
- (__mmask8)__builtin_ia32_cmpq512_mask((__v8di)(__m512i)(a), \
5009
- (__v8di)(__m512i)(b), (int)(p), \
5010
- (__mmask8)(m))
5012
+ ((__mmask8)__builtin_ia32_cmpq512_mask((__v8di)(__m512i)(a), \
5013
+ (__v8di)(__m512i)(b), (int)(p), \
5014
+ (__mmask8)(m)))
5011
5015
 
5012
5016
  #define _mm512_mask_cmp_epu64_mask(m, a, b, p) \
5013
- (__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)(__m512i)(a), \
5014
- (__v8di)(__m512i)(b), (int)(p), \
5015
- (__mmask8)(m))
5017
+ ((__mmask8)__builtin_ia32_ucmpq512_mask((__v8di)(__m512i)(a), \
5018
+ (__v8di)(__m512i)(b), (int)(p), \
5019
+ (__mmask8)(m)))
5016
5020
 
5017
5021
  #define _mm512_rol_epi32(a, b) \
5018
- (__m512i)__builtin_ia32_prold512((__v16si)(__m512i)(a), (int)(b))
5022
+ ((__m512i)__builtin_ia32_prold512((__v16si)(__m512i)(a), (int)(b)))
5019
5023
 
5020
5024
  #define _mm512_mask_rol_epi32(W, U, a, b) \
5021
- (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
5022
- (__v16si)_mm512_rol_epi32((a), (b)), \
5023
- (__v16si)(__m512i)(W))
5025
+ ((__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
5026
+ (__v16si)_mm512_rol_epi32((a), (b)), \
5027
+ (__v16si)(__m512i)(W)))
5024
5028
 
5025
5029
  #define _mm512_maskz_rol_epi32(U, a, b) \
5026
- (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
5027
- (__v16si)_mm512_rol_epi32((a), (b)), \
5028
- (__v16si)_mm512_setzero_si512())
5030
+ ((__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
5031
+ (__v16si)_mm512_rol_epi32((a), (b)), \
5032
+ (__v16si)_mm512_setzero_si512()))
5029
5033
 
5030
5034
  #define _mm512_rol_epi64(a, b) \
5031
- (__m512i)__builtin_ia32_prolq512((__v8di)(__m512i)(a), (int)(b))
5035
+ ((__m512i)__builtin_ia32_prolq512((__v8di)(__m512i)(a), (int)(b)))
5032
5036
 
5033
5037
  #define _mm512_mask_rol_epi64(W, U, a, b) \
5034
- (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
5035
- (__v8di)_mm512_rol_epi64((a), (b)), \
5036
- (__v8di)(__m512i)(W))
5038
+ ((__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
5039
+ (__v8di)_mm512_rol_epi64((a), (b)), \
5040
+ (__v8di)(__m512i)(W)))
5037
5041
 
5038
5042
  #define _mm512_maskz_rol_epi64(U, a, b) \
5039
- (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
5040
- (__v8di)_mm512_rol_epi64((a), (b)), \
5041
- (__v8di)_mm512_setzero_si512())
5043
+ ((__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
5044
+ (__v8di)_mm512_rol_epi64((a), (b)), \
5045
+ (__v8di)_mm512_setzero_si512()))
5042
5046
 
5043
5047
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
5044
5048
  _mm512_rolv_epi32 (__m512i __A, __m512i __B)
@@ -5085,30 +5089,30 @@ _mm512_maskz_rolv_epi64 (__mmask8 __U, __m512i __A, __m512i __B)
5085
5089
  }
5086
5090
 
5087
5091
  #define _mm512_ror_epi32(A, B) \
5088
- (__m512i)__builtin_ia32_prord512((__v16si)(__m512i)(A), (int)(B))
5092
+ ((__m512i)__builtin_ia32_prord512((__v16si)(__m512i)(A), (int)(B)))
5089
5093
 
5090
5094
  #define _mm512_mask_ror_epi32(W, U, A, B) \
5091
- (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
5092
- (__v16si)_mm512_ror_epi32((A), (B)), \
5093
- (__v16si)(__m512i)(W))
5095
+ ((__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
5096
+ (__v16si)_mm512_ror_epi32((A), (B)), \
5097
+ (__v16si)(__m512i)(W)))
5094
5098
 
5095
5099
  #define _mm512_maskz_ror_epi32(U, A, B) \
5096
- (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
5097
- (__v16si)_mm512_ror_epi32((A), (B)), \
5098
- (__v16si)_mm512_setzero_si512())
5100
+ ((__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
5101
+ (__v16si)_mm512_ror_epi32((A), (B)), \
5102
+ (__v16si)_mm512_setzero_si512()))
5099
5103
 
5100
5104
  #define _mm512_ror_epi64(A, B) \
5101
- (__m512i)__builtin_ia32_prorq512((__v8di)(__m512i)(A), (int)(B))
5105
+ ((__m512i)__builtin_ia32_prorq512((__v8di)(__m512i)(A), (int)(B)))
5102
5106
 
5103
5107
  #define _mm512_mask_ror_epi64(W, U, A, B) \
5104
- (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
5105
- (__v8di)_mm512_ror_epi64((A), (B)), \
5106
- (__v8di)(__m512i)(W))
5108
+ ((__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
5109
+ (__v8di)_mm512_ror_epi64((A), (B)), \
5110
+ (__v8di)(__m512i)(W)))
5107
5111
 
5108
5112
  #define _mm512_maskz_ror_epi64(U, A, B) \
5109
- (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
5110
- (__v8di)_mm512_ror_epi64((A), (B)), \
5111
- (__v8di)_mm512_setzero_si512())
5113
+ ((__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
5114
+ (__v8di)_mm512_ror_epi64((A), (B)), \
5115
+ (__v8di)_mm512_setzero_si512()))
5112
5116
 
5113
5117
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
5114
5118
  _mm512_slli_epi32(__m512i __A, unsigned int __B)
@@ -5304,168 +5308,168 @@ _mm512_maskz_movedup_pd (__mmask8 __U, __m512d __A)
5304
5308
  }
5305
5309
 
5306
5310
  #define _mm512_fixupimm_round_pd(A, B, C, imm, R) \
5307
- (__m512d)__builtin_ia32_fixupimmpd512_mask((__v8df)(__m512d)(A), \
5308
- (__v8df)(__m512d)(B), \
5309
- (__v8di)(__m512i)(C), (int)(imm), \
5310
- (__mmask8)-1, (int)(R))
5311
+ ((__m512d)__builtin_ia32_fixupimmpd512_mask((__v8df)(__m512d)(A), \
5312
+ (__v8df)(__m512d)(B), \
5313
+ (__v8di)(__m512i)(C), (int)(imm), \
5314
+ (__mmask8)-1, (int)(R)))
5311
5315
 
5312
5316
  #define _mm512_mask_fixupimm_round_pd(A, U, B, C, imm, R) \
5313
- (__m512d)__builtin_ia32_fixupimmpd512_mask((__v8df)(__m512d)(A), \
5314
- (__v8df)(__m512d)(B), \
5315
- (__v8di)(__m512i)(C), (int)(imm), \
5316
- (__mmask8)(U), (int)(R))
5317
+ ((__m512d)__builtin_ia32_fixupimmpd512_mask((__v8df)(__m512d)(A), \
5318
+ (__v8df)(__m512d)(B), \
5319
+ (__v8di)(__m512i)(C), (int)(imm), \
5320
+ (__mmask8)(U), (int)(R)))
5317
5321
 
5318
5322
  #define _mm512_fixupimm_pd(A, B, C, imm) \
5319
- (__m512d)__builtin_ia32_fixupimmpd512_mask((__v8df)(__m512d)(A), \
5320
- (__v8df)(__m512d)(B), \
5321
- (__v8di)(__m512i)(C), (int)(imm), \
5322
- (__mmask8)-1, \
5323
- _MM_FROUND_CUR_DIRECTION)
5323
+ ((__m512d)__builtin_ia32_fixupimmpd512_mask((__v8df)(__m512d)(A), \
5324
+ (__v8df)(__m512d)(B), \
5325
+ (__v8di)(__m512i)(C), (int)(imm), \
5326
+ (__mmask8)-1, \
5327
+ _MM_FROUND_CUR_DIRECTION))
5324
5328
 
5325
5329
  #define _mm512_mask_fixupimm_pd(A, U, B, C, imm) \
5326
- (__m512d)__builtin_ia32_fixupimmpd512_mask((__v8df)(__m512d)(A), \
5327
- (__v8df)(__m512d)(B), \
5328
- (__v8di)(__m512i)(C), (int)(imm), \
5329
- (__mmask8)(U), \
5330
- _MM_FROUND_CUR_DIRECTION)
5330
+ ((__m512d)__builtin_ia32_fixupimmpd512_mask((__v8df)(__m512d)(A), \
5331
+ (__v8df)(__m512d)(B), \
5332
+ (__v8di)(__m512i)(C), (int)(imm), \
5333
+ (__mmask8)(U), \
5334
+ _MM_FROUND_CUR_DIRECTION))
5331
5335
 
5332
5336
  #define _mm512_maskz_fixupimm_round_pd(U, A, B, C, imm, R) \
5333
- (__m512d)__builtin_ia32_fixupimmpd512_maskz((__v8df)(__m512d)(A), \
5334
- (__v8df)(__m512d)(B), \
5335
- (__v8di)(__m512i)(C), \
5336
- (int)(imm), (__mmask8)(U), \
5337
- (int)(R))
5337
+ ((__m512d)__builtin_ia32_fixupimmpd512_maskz((__v8df)(__m512d)(A), \
5338
+ (__v8df)(__m512d)(B), \
5339
+ (__v8di)(__m512i)(C), \
5340
+ (int)(imm), (__mmask8)(U), \
5341
+ (int)(R)))
5338
5342
 
5339
5343
  #define _mm512_maskz_fixupimm_pd(U, A, B, C, imm) \
5340
- (__m512d)__builtin_ia32_fixupimmpd512_maskz((__v8df)(__m512d)(A), \
5341
- (__v8df)(__m512d)(B), \
5342
- (__v8di)(__m512i)(C), \
5343
- (int)(imm), (__mmask8)(U), \
5344
- _MM_FROUND_CUR_DIRECTION)
5344
+ ((__m512d)__builtin_ia32_fixupimmpd512_maskz((__v8df)(__m512d)(A), \
5345
+ (__v8df)(__m512d)(B), \
5346
+ (__v8di)(__m512i)(C), \
5347
+ (int)(imm), (__mmask8)(U), \
5348
+ _MM_FROUND_CUR_DIRECTION))
5345
5349
 
5346
5350
  #define _mm512_fixupimm_round_ps(A, B, C, imm, R) \
5347
- (__m512)__builtin_ia32_fixupimmps512_mask((__v16sf)(__m512)(A), \
5348
- (__v16sf)(__m512)(B), \
5349
- (__v16si)(__m512i)(C), (int)(imm), \
5350
- (__mmask16)-1, (int)(R))
5351
+ ((__m512)__builtin_ia32_fixupimmps512_mask((__v16sf)(__m512)(A), \
5352
+ (__v16sf)(__m512)(B), \
5353
+ (__v16si)(__m512i)(C), (int)(imm), \
5354
+ (__mmask16)-1, (int)(R)))
5351
5355
 
5352
5356
  #define _mm512_mask_fixupimm_round_ps(A, U, B, C, imm, R) \
5353
- (__m512)__builtin_ia32_fixupimmps512_mask((__v16sf)(__m512)(A), \
5354
- (__v16sf)(__m512)(B), \
5355
- (__v16si)(__m512i)(C), (int)(imm), \
5356
- (__mmask16)(U), (int)(R))
5357
+ ((__m512)__builtin_ia32_fixupimmps512_mask((__v16sf)(__m512)(A), \
5358
+ (__v16sf)(__m512)(B), \
5359
+ (__v16si)(__m512i)(C), (int)(imm), \
5360
+ (__mmask16)(U), (int)(R)))
5357
5361
 
5358
5362
  #define _mm512_fixupimm_ps(A, B, C, imm) \
5359
- (__m512)__builtin_ia32_fixupimmps512_mask((__v16sf)(__m512)(A), \
5360
- (__v16sf)(__m512)(B), \
5361
- (__v16si)(__m512i)(C), (int)(imm), \
5362
- (__mmask16)-1, \
5363
- _MM_FROUND_CUR_DIRECTION)
5363
+ ((__m512)__builtin_ia32_fixupimmps512_mask((__v16sf)(__m512)(A), \
5364
+ (__v16sf)(__m512)(B), \
5365
+ (__v16si)(__m512i)(C), (int)(imm), \
5366
+ (__mmask16)-1, \
5367
+ _MM_FROUND_CUR_DIRECTION))
5364
5368
 
5365
5369
  #define _mm512_mask_fixupimm_ps(A, U, B, C, imm) \
5366
- (__m512)__builtin_ia32_fixupimmps512_mask((__v16sf)(__m512)(A), \
5367
- (__v16sf)(__m512)(B), \
5368
- (__v16si)(__m512i)(C), (int)(imm), \
5369
- (__mmask16)(U), \
5370
- _MM_FROUND_CUR_DIRECTION)
5370
+ ((__m512)__builtin_ia32_fixupimmps512_mask((__v16sf)(__m512)(A), \
5371
+ (__v16sf)(__m512)(B), \
5372
+ (__v16si)(__m512i)(C), (int)(imm), \
5373
+ (__mmask16)(U), \
5374
+ _MM_FROUND_CUR_DIRECTION))
5371
5375
 
5372
5376
  #define _mm512_maskz_fixupimm_round_ps(U, A, B, C, imm, R) \
5373
- (__m512)__builtin_ia32_fixupimmps512_maskz((__v16sf)(__m512)(A), \
5374
- (__v16sf)(__m512)(B), \
5375
- (__v16si)(__m512i)(C), \
5376
- (int)(imm), (__mmask16)(U), \
5377
- (int)(R))
5377
+ ((__m512)__builtin_ia32_fixupimmps512_maskz((__v16sf)(__m512)(A), \
5378
+ (__v16sf)(__m512)(B), \
5379
+ (__v16si)(__m512i)(C), \
5380
+ (int)(imm), (__mmask16)(U), \
5381
+ (int)(R)))
5378
5382
 
5379
5383
  #define _mm512_maskz_fixupimm_ps(U, A, B, C, imm) \
5380
- (__m512)__builtin_ia32_fixupimmps512_maskz((__v16sf)(__m512)(A), \
5381
- (__v16sf)(__m512)(B), \
5382
- (__v16si)(__m512i)(C), \
5383
- (int)(imm), (__mmask16)(U), \
5384
- _MM_FROUND_CUR_DIRECTION)
5384
+ ((__m512)__builtin_ia32_fixupimmps512_maskz((__v16sf)(__m512)(A), \
5385
+ (__v16sf)(__m512)(B), \
5386
+ (__v16si)(__m512i)(C), \
5387
+ (int)(imm), (__mmask16)(U), \
5388
+ _MM_FROUND_CUR_DIRECTION))
5385
5389
 
5386
5390
  #define _mm_fixupimm_round_sd(A, B, C, imm, R) \
5387
- (__m128d)__builtin_ia32_fixupimmsd_mask((__v2df)(__m128d)(A), \
5388
- (__v2df)(__m128d)(B), \
5389
- (__v2di)(__m128i)(C), (int)(imm), \
5390
- (__mmask8)-1, (int)(R))
5391
+ ((__m128d)__builtin_ia32_fixupimmsd_mask((__v2df)(__m128d)(A), \
5392
+ (__v2df)(__m128d)(B), \
5393
+ (__v2di)(__m128i)(C), (int)(imm), \
5394
+ (__mmask8)-1, (int)(R)))
5391
5395
 
5392
5396
  #define _mm_mask_fixupimm_round_sd(A, U, B, C, imm, R) \
5393
- (__m128d)__builtin_ia32_fixupimmsd_mask((__v2df)(__m128d)(A), \
5394
- (__v2df)(__m128d)(B), \
5395
- (__v2di)(__m128i)(C), (int)(imm), \
5396
- (__mmask8)(U), (int)(R))
5397
+ ((__m128d)__builtin_ia32_fixupimmsd_mask((__v2df)(__m128d)(A), \
5398
+ (__v2df)(__m128d)(B), \
5399
+ (__v2di)(__m128i)(C), (int)(imm), \
5400
+ (__mmask8)(U), (int)(R)))
5397
5401
 
5398
5402
  #define _mm_fixupimm_sd(A, B, C, imm) \
5399
- (__m128d)__builtin_ia32_fixupimmsd_mask((__v2df)(__m128d)(A), \
5400
- (__v2df)(__m128d)(B), \
5401
- (__v2di)(__m128i)(C), (int)(imm), \
5402
- (__mmask8)-1, \
5403
- _MM_FROUND_CUR_DIRECTION)
5404
-
5405
- #define _mm_mask_fixupimm_sd(A, U, B, C, imm) \
5406
- (__m128d)__builtin_ia32_fixupimmsd_mask((__v2df)(__m128d)(A), \
5407
- (__v2df)(__m128d)(B), \
5408
- (__v2di)(__m128i)(C), (int)(imm), \
5409
- (__mmask8)(U), \
5410
- _MM_FROUND_CUR_DIRECTION)
5411
-
5412
- #define _mm_maskz_fixupimm_round_sd(U, A, B, C, imm, R) \
5413
- (__m128d)__builtin_ia32_fixupimmsd_maskz((__v2df)(__m128d)(A), \
5403
+ ((__m128d)__builtin_ia32_fixupimmsd_mask((__v2df)(__m128d)(A), \
5414
5404
  (__v2df)(__m128d)(B), \
5415
5405
  (__v2di)(__m128i)(C), (int)(imm), \
5416
- (__mmask8)(U), (int)(R))
5406
+ (__mmask8)-1, \
5407
+ _MM_FROUND_CUR_DIRECTION))
5417
5408
 
5418
- #define _mm_maskz_fixupimm_sd(U, A, B, C, imm) \
5419
- (__m128d)__builtin_ia32_fixupimmsd_maskz((__v2df)(__m128d)(A), \
5409
+ #define _mm_mask_fixupimm_sd(A, U, B, C, imm) \
5410
+ ((__m128d)__builtin_ia32_fixupimmsd_mask((__v2df)(__m128d)(A), \
5420
5411
  (__v2df)(__m128d)(B), \
5421
5412
  (__v2di)(__m128i)(C), (int)(imm), \
5422
5413
  (__mmask8)(U), \
5423
- _MM_FROUND_CUR_DIRECTION)
5414
+ _MM_FROUND_CUR_DIRECTION))
5415
+
5416
+ #define _mm_maskz_fixupimm_round_sd(U, A, B, C, imm, R) \
5417
+ ((__m128d)__builtin_ia32_fixupimmsd_maskz((__v2df)(__m128d)(A), \
5418
+ (__v2df)(__m128d)(B), \
5419
+ (__v2di)(__m128i)(C), (int)(imm), \
5420
+ (__mmask8)(U), (int)(R)))
5421
+
5422
+ #define _mm_maskz_fixupimm_sd(U, A, B, C, imm) \
5423
+ ((__m128d)__builtin_ia32_fixupimmsd_maskz((__v2df)(__m128d)(A), \
5424
+ (__v2df)(__m128d)(B), \
5425
+ (__v2di)(__m128i)(C), (int)(imm), \
5426
+ (__mmask8)(U), \
5427
+ _MM_FROUND_CUR_DIRECTION))
5424
5428
 
5425
5429
  #define _mm_fixupimm_round_ss(A, B, C, imm, R) \
5426
- (__m128)__builtin_ia32_fixupimmss_mask((__v4sf)(__m128)(A), \
5427
- (__v4sf)(__m128)(B), \
5428
- (__v4si)(__m128i)(C), (int)(imm), \
5429
- (__mmask8)-1, (int)(R))
5430
+ ((__m128)__builtin_ia32_fixupimmss_mask((__v4sf)(__m128)(A), \
5431
+ (__v4sf)(__m128)(B), \
5432
+ (__v4si)(__m128i)(C), (int)(imm), \
5433
+ (__mmask8)-1, (int)(R)))
5430
5434
 
5431
5435
  #define _mm_mask_fixupimm_round_ss(A, U, B, C, imm, R) \
5432
- (__m128)__builtin_ia32_fixupimmss_mask((__v4sf)(__m128)(A), \
5433
- (__v4sf)(__m128)(B), \
5434
- (__v4si)(__m128i)(C), (int)(imm), \
5435
- (__mmask8)(U), (int)(R))
5436
+ ((__m128)__builtin_ia32_fixupimmss_mask((__v4sf)(__m128)(A), \
5437
+ (__v4sf)(__m128)(B), \
5438
+ (__v4si)(__m128i)(C), (int)(imm), \
5439
+ (__mmask8)(U), (int)(R)))
5436
5440
 
5437
5441
  #define _mm_fixupimm_ss(A, B, C, imm) \
5438
- (__m128)__builtin_ia32_fixupimmss_mask((__v4sf)(__m128)(A), \
5439
- (__v4sf)(__m128)(B), \
5440
- (__v4si)(__m128i)(C), (int)(imm), \
5441
- (__mmask8)-1, \
5442
- _MM_FROUND_CUR_DIRECTION)
5443
-
5444
- #define _mm_mask_fixupimm_ss(A, U, B, C, imm) \
5445
- (__m128)__builtin_ia32_fixupimmss_mask((__v4sf)(__m128)(A), \
5446
- (__v4sf)(__m128)(B), \
5447
- (__v4si)(__m128i)(C), (int)(imm), \
5448
- (__mmask8)(U), \
5449
- _MM_FROUND_CUR_DIRECTION)
5450
-
5451
- #define _mm_maskz_fixupimm_round_ss(U, A, B, C, imm, R) \
5452
- (__m128)__builtin_ia32_fixupimmss_maskz((__v4sf)(__m128)(A), \
5442
+ ((__m128)__builtin_ia32_fixupimmss_mask((__v4sf)(__m128)(A), \
5453
5443
  (__v4sf)(__m128)(B), \
5454
5444
  (__v4si)(__m128i)(C), (int)(imm), \
5455
- (__mmask8)(U), (int)(R))
5445
+ (__mmask8)-1, \
5446
+ _MM_FROUND_CUR_DIRECTION))
5456
5447
 
5457
- #define _mm_maskz_fixupimm_ss(U, A, B, C, imm) \
5458
- (__m128)__builtin_ia32_fixupimmss_maskz((__v4sf)(__m128)(A), \
5448
+ #define _mm_mask_fixupimm_ss(A, U, B, C, imm) \
5449
+ ((__m128)__builtin_ia32_fixupimmss_mask((__v4sf)(__m128)(A), \
5459
5450
  (__v4sf)(__m128)(B), \
5460
5451
  (__v4si)(__m128i)(C), (int)(imm), \
5461
5452
  (__mmask8)(U), \
5462
- _MM_FROUND_CUR_DIRECTION)
5453
+ _MM_FROUND_CUR_DIRECTION))
5454
+
5455
+ #define _mm_maskz_fixupimm_round_ss(U, A, B, C, imm, R) \
5456
+ ((__m128)__builtin_ia32_fixupimmss_maskz((__v4sf)(__m128)(A), \
5457
+ (__v4sf)(__m128)(B), \
5458
+ (__v4si)(__m128i)(C), (int)(imm), \
5459
+ (__mmask8)(U), (int)(R)))
5460
+
5461
+ #define _mm_maskz_fixupimm_ss(U, A, B, C, imm) \
5462
+ ((__m128)__builtin_ia32_fixupimmss_maskz((__v4sf)(__m128)(A), \
5463
+ (__v4sf)(__m128)(B), \
5464
+ (__v4si)(__m128i)(C), (int)(imm), \
5465
+ (__mmask8)(U), \
5466
+ _MM_FROUND_CUR_DIRECTION))
5463
5467
 
5464
5468
  #define _mm_getexp_round_sd(A, B, R) \
5465
- (__m128d)__builtin_ia32_getexpsd128_round_mask((__v2df)(__m128d)(A), \
5466
- (__v2df)(__m128d)(B), \
5467
- (__v2df)_mm_setzero_pd(), \
5468
- (__mmask8)-1, (int)(R))
5469
+ ((__m128d)__builtin_ia32_getexpsd128_round_mask((__v2df)(__m128d)(A), \
5470
+ (__v2df)(__m128d)(B), \
5471
+ (__v2df)_mm_setzero_pd(), \
5472
+ (__mmask8)-1, (int)(R)))
5469
5473
 
5470
5474
 
5471
5475
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
@@ -5486,10 +5490,10 @@ _mm_mask_getexp_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
5486
5490
  }
5487
5491
 
5488
5492
  #define _mm_mask_getexp_round_sd(W, U, A, B, R) \
5489
- (__m128d)__builtin_ia32_getexpsd128_round_mask((__v2df)(__m128d)(A), \
5490
- (__v2df)(__m128d)(B), \
5491
- (__v2df)(__m128d)(W), \
5492
- (__mmask8)(U), (int)(R))
5493
+ ((__m128d)__builtin_ia32_getexpsd128_round_mask((__v2df)(__m128d)(A), \
5494
+ (__v2df)(__m128d)(B), \
5495
+ (__v2df)(__m128d)(W), \
5496
+ (__mmask8)(U), (int)(R)))
5493
5497
 
5494
5498
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
5495
5499
  _mm_maskz_getexp_sd (__mmask8 __U, __m128d __A, __m128d __B)
@@ -5502,16 +5506,16 @@ _mm_maskz_getexp_sd (__mmask8 __U, __m128d __A, __m128d __B)
5502
5506
  }
5503
5507
 
5504
5508
  #define _mm_maskz_getexp_round_sd(U, A, B, R) \
5505
- (__m128d)__builtin_ia32_getexpsd128_round_mask((__v2df)(__m128d)(A), \
5506
- (__v2df)(__m128d)(B), \
5507
- (__v2df)_mm_setzero_pd(), \
5508
- (__mmask8)(U), (int)(R))
5509
+ ((__m128d)__builtin_ia32_getexpsd128_round_mask((__v2df)(__m128d)(A), \
5510
+ (__v2df)(__m128d)(B), \
5511
+ (__v2df)_mm_setzero_pd(), \
5512
+ (__mmask8)(U), (int)(R)))
5509
5513
 
5510
5514
  #define _mm_getexp_round_ss(A, B, R) \
5511
- (__m128)__builtin_ia32_getexpss128_round_mask((__v4sf)(__m128)(A), \
5512
- (__v4sf)(__m128)(B), \
5513
- (__v4sf)_mm_setzero_ps(), \
5514
- (__mmask8)-1, (int)(R))
5515
+ ((__m128)__builtin_ia32_getexpss128_round_mask((__v4sf)(__m128)(A), \
5516
+ (__v4sf)(__m128)(B), \
5517
+ (__v4sf)_mm_setzero_ps(), \
5518
+ (__mmask8)-1, (int)(R)))
5515
5519
 
5516
5520
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
5517
5521
  _mm_getexp_ss (__m128 __A, __m128 __B)
@@ -5531,10 +5535,10 @@ _mm_mask_getexp_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
5531
5535
  }
5532
5536
 
5533
5537
  #define _mm_mask_getexp_round_ss(W, U, A, B, R) \
5534
- (__m128)__builtin_ia32_getexpss128_round_mask((__v4sf)(__m128)(A), \
5535
- (__v4sf)(__m128)(B), \
5536
- (__v4sf)(__m128)(W), \
5537
- (__mmask8)(U), (int)(R))
5538
+ ((__m128)__builtin_ia32_getexpss128_round_mask((__v4sf)(__m128)(A), \
5539
+ (__v4sf)(__m128)(B), \
5540
+ (__v4sf)(__m128)(W), \
5541
+ (__mmask8)(U), (int)(R)))
5538
5542
 
5539
5543
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
5540
5544
  _mm_maskz_getexp_ss (__mmask8 __U, __m128 __A, __m128 __B)
@@ -5547,100 +5551,100 @@ _mm_maskz_getexp_ss (__mmask8 __U, __m128 __A, __m128 __B)
5547
5551
  }
5548
5552
 
5549
5553
  #define _mm_maskz_getexp_round_ss(U, A, B, R) \
5550
- (__m128)__builtin_ia32_getexpss128_round_mask((__v4sf)(__m128)(A), \
5551
- (__v4sf)(__m128)(B), \
5552
- (__v4sf)_mm_setzero_ps(), \
5553
- (__mmask8)(U), (int)(R))
5554
+ ((__m128)__builtin_ia32_getexpss128_round_mask((__v4sf)(__m128)(A), \
5555
+ (__v4sf)(__m128)(B), \
5556
+ (__v4sf)_mm_setzero_ps(), \
5557
+ (__mmask8)(U), (int)(R)))
5554
5558
 
5555
5559
  #define _mm_getmant_round_sd(A, B, C, D, R) \
5556
- (__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
5557
- (__v2df)(__m128d)(B), \
5558
- (int)(((D)<<2) | (C)), \
5559
- (__v2df)_mm_setzero_pd(), \
5560
- (__mmask8)-1, (int)(R))
5560
+ ((__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
5561
+ (__v2df)(__m128d)(B), \
5562
+ (int)(((D)<<2) | (C)), \
5563
+ (__v2df)_mm_setzero_pd(), \
5564
+ (__mmask8)-1, (int)(R)))
5561
5565
 
5562
5566
  #define _mm_getmant_sd(A, B, C, D) \
5563
- (__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
5564
- (__v2df)(__m128d)(B), \
5565
- (int)(((D)<<2) | (C)), \
5566
- (__v2df)_mm_setzero_pd(), \
5567
- (__mmask8)-1, \
5568
- _MM_FROUND_CUR_DIRECTION)
5567
+ ((__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
5568
+ (__v2df)(__m128d)(B), \
5569
+ (int)(((D)<<2) | (C)), \
5570
+ (__v2df)_mm_setzero_pd(), \
5571
+ (__mmask8)-1, \
5572
+ _MM_FROUND_CUR_DIRECTION))
5569
5573
 
5570
5574
  #define _mm_mask_getmant_sd(W, U, A, B, C, D) \
5571
- (__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
5572
- (__v2df)(__m128d)(B), \
5573
- (int)(((D)<<2) | (C)), \
5574
- (__v2df)(__m128d)(W), \
5575
- (__mmask8)(U), \
5576
- _MM_FROUND_CUR_DIRECTION)
5575
+ ((__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
5576
+ (__v2df)(__m128d)(B), \
5577
+ (int)(((D)<<2) | (C)), \
5578
+ (__v2df)(__m128d)(W), \
5579
+ (__mmask8)(U), \
5580
+ _MM_FROUND_CUR_DIRECTION))
5577
5581
 
5578
5582
  #define _mm_mask_getmant_round_sd(W, U, A, B, C, D, R) \
5579
- (__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
5580
- (__v2df)(__m128d)(B), \
5581
- (int)(((D)<<2) | (C)), \
5582
- (__v2df)(__m128d)(W), \
5583
- (__mmask8)(U), (int)(R))
5583
+ ((__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
5584
+ (__v2df)(__m128d)(B), \
5585
+ (int)(((D)<<2) | (C)), \
5586
+ (__v2df)(__m128d)(W), \
5587
+ (__mmask8)(U), (int)(R)))
5584
5588
 
5585
5589
  #define _mm_maskz_getmant_sd(U, A, B, C, D) \
5586
- (__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
5587
- (__v2df)(__m128d)(B), \
5588
- (int)(((D)<<2) | (C)), \
5589
- (__v2df)_mm_setzero_pd(), \
5590
- (__mmask8)(U), \
5591
- _MM_FROUND_CUR_DIRECTION)
5590
+ ((__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
5591
+ (__v2df)(__m128d)(B), \
5592
+ (int)(((D)<<2) | (C)), \
5593
+ (__v2df)_mm_setzero_pd(), \
5594
+ (__mmask8)(U), \
5595
+ _MM_FROUND_CUR_DIRECTION))
5592
5596
 
5593
5597
  #define _mm_maskz_getmant_round_sd(U, A, B, C, D, R) \
5594
- (__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
5595
- (__v2df)(__m128d)(B), \
5596
- (int)(((D)<<2) | (C)), \
5597
- (__v2df)_mm_setzero_pd(), \
5598
- (__mmask8)(U), (int)(R))
5598
+ ((__m128d)__builtin_ia32_getmantsd_round_mask((__v2df)(__m128d)(A), \
5599
+ (__v2df)(__m128d)(B), \
5600
+ (int)(((D)<<2) | (C)), \
5601
+ (__v2df)_mm_setzero_pd(), \
5602
+ (__mmask8)(U), (int)(R)))
5599
5603
 
5600
5604
  #define _mm_getmant_round_ss(A, B, C, D, R) \
5601
- (__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
5602
- (__v4sf)(__m128)(B), \
5603
- (int)(((D)<<2) | (C)), \
5604
- (__v4sf)_mm_setzero_ps(), \
5605
- (__mmask8)-1, (int)(R))
5605
+ ((__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
5606
+ (__v4sf)(__m128)(B), \
5607
+ (int)(((D)<<2) | (C)), \
5608
+ (__v4sf)_mm_setzero_ps(), \
5609
+ (__mmask8)-1, (int)(R)))
5606
5610
 
5607
5611
  #define _mm_getmant_ss(A, B, C, D) \
5608
- (__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
5609
- (__v4sf)(__m128)(B), \
5610
- (int)(((D)<<2) | (C)), \
5611
- (__v4sf)_mm_setzero_ps(), \
5612
- (__mmask8)-1, \
5613
- _MM_FROUND_CUR_DIRECTION)
5612
+ ((__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
5613
+ (__v4sf)(__m128)(B), \
5614
+ (int)(((D)<<2) | (C)), \
5615
+ (__v4sf)_mm_setzero_ps(), \
5616
+ (__mmask8)-1, \
5617
+ _MM_FROUND_CUR_DIRECTION))
5614
5618
 
5615
5619
  #define _mm_mask_getmant_ss(W, U, A, B, C, D) \
5616
- (__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
5617
- (__v4sf)(__m128)(B), \
5618
- (int)(((D)<<2) | (C)), \
5619
- (__v4sf)(__m128)(W), \
5620
- (__mmask8)(U), \
5621
- _MM_FROUND_CUR_DIRECTION)
5620
+ ((__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
5621
+ (__v4sf)(__m128)(B), \
5622
+ (int)(((D)<<2) | (C)), \
5623
+ (__v4sf)(__m128)(W), \
5624
+ (__mmask8)(U), \
5625
+ _MM_FROUND_CUR_DIRECTION))
5622
5626
 
5623
5627
  #define _mm_mask_getmant_round_ss(W, U, A, B, C, D, R) \
5624
- (__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
5625
- (__v4sf)(__m128)(B), \
5626
- (int)(((D)<<2) | (C)), \
5627
- (__v4sf)(__m128)(W), \
5628
- (__mmask8)(U), (int)(R))
5628
+ ((__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
5629
+ (__v4sf)(__m128)(B), \
5630
+ (int)(((D)<<2) | (C)), \
5631
+ (__v4sf)(__m128)(W), \
5632
+ (__mmask8)(U), (int)(R)))
5629
5633
 
5630
5634
  #define _mm_maskz_getmant_ss(U, A, B, C, D) \
5631
- (__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
5632
- (__v4sf)(__m128)(B), \
5633
- (int)(((D)<<2) | (C)), \
5634
- (__v4sf)_mm_setzero_ps(), \
5635
- (__mmask8)(U), \
5636
- _MM_FROUND_CUR_DIRECTION)
5635
+ ((__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
5636
+ (__v4sf)(__m128)(B), \
5637
+ (int)(((D)<<2) | (C)), \
5638
+ (__v4sf)_mm_setzero_ps(), \
5639
+ (__mmask8)(U), \
5640
+ _MM_FROUND_CUR_DIRECTION))
5637
5641
 
5638
5642
  #define _mm_maskz_getmant_round_ss(U, A, B, C, D, R) \
5639
- (__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
5640
- (__v4sf)(__m128)(B), \
5641
- (int)(((D)<<2) | (C)), \
5642
- (__v4sf)_mm_setzero_ps(), \
5643
- (__mmask8)(U), (int)(R))
5643
+ ((__m128)__builtin_ia32_getmantss_round_mask((__v4sf)(__m128)(A), \
5644
+ (__v4sf)(__m128)(B), \
5645
+ (int)(((D)<<2) | (C)), \
5646
+ (__v4sf)_mm_setzero_ps(), \
5647
+ (__mmask8)(U), (int)(R)))
5644
5648
 
5645
5649
  static __inline__ __mmask16 __DEFAULT_FN_ATTRS
5646
5650
  _mm512_kmov (__mmask16 __A)
@@ -5649,16 +5653,16 @@ _mm512_kmov (__mmask16 __A)
5649
5653
  }
5650
5654
 
5651
5655
  #define _mm_comi_round_sd(A, B, P, R) \
5652
- (int)__builtin_ia32_vcomisd((__v2df)(__m128d)(A), (__v2df)(__m128d)(B), \
5653
- (int)(P), (int)(R))
5656
+ ((int)__builtin_ia32_vcomisd((__v2df)(__m128d)(A), (__v2df)(__m128d)(B), \
5657
+ (int)(P), (int)(R)))
5654
5658
 
5655
5659
  #define _mm_comi_round_ss(A, B, P, R) \
5656
- (int)__builtin_ia32_vcomiss((__v4sf)(__m128)(A), (__v4sf)(__m128)(B), \
5657
- (int)(P), (int)(R))
5660
+ ((int)__builtin_ia32_vcomiss((__v4sf)(__m128)(A), (__v4sf)(__m128)(B), \
5661
+ (int)(P), (int)(R)))
5658
5662
 
5659
5663
  #ifdef __x86_64__
5660
5664
  #define _mm_cvt_roundsd_si64(A, R) \
5661
- (long long)__builtin_ia32_vcvtsd2si64((__v2df)(__m128d)(A), (int)(R))
5665
+ ((long long)__builtin_ia32_vcvtsd2si64((__v2df)(__m128d)(A), (int)(R)))
5662
5666
  #endif
5663
5667
 
5664
5668
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
@@ -5926,54 +5930,54 @@ _mm512_maskz_srlv_epi64(__mmask8 __U, __m512i __X, __m512i __Y)
5926
5930
  }
5927
5931
 
5928
5932
  #define _mm512_ternarylogic_epi32(A, B, C, imm) \
5929
- (__m512i)__builtin_ia32_pternlogd512_mask((__v16si)(__m512i)(A), \
5930
- (__v16si)(__m512i)(B), \
5931
- (__v16si)(__m512i)(C), (int)(imm), \
5932
- (__mmask16)-1)
5933
+ ((__m512i)__builtin_ia32_pternlogd512_mask((__v16si)(__m512i)(A), \
5934
+ (__v16si)(__m512i)(B), \
5935
+ (__v16si)(__m512i)(C), (int)(imm), \
5936
+ (__mmask16)-1))
5933
5937
 
5934
5938
  #define _mm512_mask_ternarylogic_epi32(A, U, B, C, imm) \
5935
- (__m512i)__builtin_ia32_pternlogd512_mask((__v16si)(__m512i)(A), \
5936
- (__v16si)(__m512i)(B), \
5937
- (__v16si)(__m512i)(C), (int)(imm), \
5938
- (__mmask16)(U))
5939
+ ((__m512i)__builtin_ia32_pternlogd512_mask((__v16si)(__m512i)(A), \
5940
+ (__v16si)(__m512i)(B), \
5941
+ (__v16si)(__m512i)(C), (int)(imm), \
5942
+ (__mmask16)(U)))
5939
5943
 
5940
5944
  #define _mm512_maskz_ternarylogic_epi32(U, A, B, C, imm) \
5941
- (__m512i)__builtin_ia32_pternlogd512_maskz((__v16si)(__m512i)(A), \
5942
- (__v16si)(__m512i)(B), \
5943
- (__v16si)(__m512i)(C), \
5944
- (int)(imm), (__mmask16)(U))
5945
+ ((__m512i)__builtin_ia32_pternlogd512_maskz((__v16si)(__m512i)(A), \
5946
+ (__v16si)(__m512i)(B), \
5947
+ (__v16si)(__m512i)(C), \
5948
+ (int)(imm), (__mmask16)(U)))
5945
5949
 
5946
5950
  #define _mm512_ternarylogic_epi64(A, B, C, imm) \
5947
- (__m512i)__builtin_ia32_pternlogq512_mask((__v8di)(__m512i)(A), \
5948
- (__v8di)(__m512i)(B), \
5949
- (__v8di)(__m512i)(C), (int)(imm), \
5950
- (__mmask8)-1)
5951
+ ((__m512i)__builtin_ia32_pternlogq512_mask((__v8di)(__m512i)(A), \
5952
+ (__v8di)(__m512i)(B), \
5953
+ (__v8di)(__m512i)(C), (int)(imm), \
5954
+ (__mmask8)-1))
5951
5955
 
5952
5956
  #define _mm512_mask_ternarylogic_epi64(A, U, B, C, imm) \
5953
- (__m512i)__builtin_ia32_pternlogq512_mask((__v8di)(__m512i)(A), \
5954
- (__v8di)(__m512i)(B), \
5955
- (__v8di)(__m512i)(C), (int)(imm), \
5956
- (__mmask8)(U))
5957
-
5958
- #define _mm512_maskz_ternarylogic_epi64(U, A, B, C, imm) \
5959
- (__m512i)__builtin_ia32_pternlogq512_maskz((__v8di)(__m512i)(A), \
5957
+ ((__m512i)__builtin_ia32_pternlogq512_mask((__v8di)(__m512i)(A), \
5960
5958
  (__v8di)(__m512i)(B), \
5961
5959
  (__v8di)(__m512i)(C), (int)(imm), \
5962
- (__mmask8)(U))
5960
+ (__mmask8)(U)))
5961
+
5962
+ #define _mm512_maskz_ternarylogic_epi64(U, A, B, C, imm) \
5963
+ ((__m512i)__builtin_ia32_pternlogq512_maskz((__v8di)(__m512i)(A), \
5964
+ (__v8di)(__m512i)(B), \
5965
+ (__v8di)(__m512i)(C), (int)(imm), \
5966
+ (__mmask8)(U)))
5963
5967
 
5964
5968
  #ifdef __x86_64__
5965
5969
  #define _mm_cvt_roundsd_i64(A, R) \
5966
- (long long)__builtin_ia32_vcvtsd2si64((__v2df)(__m128d)(A), (int)(R))
5970
+ ((long long)__builtin_ia32_vcvtsd2si64((__v2df)(__m128d)(A), (int)(R)))
5967
5971
  #endif
5968
5972
 
5969
5973
  #define _mm_cvt_roundsd_si32(A, R) \
5970
- (int)__builtin_ia32_vcvtsd2si32((__v2df)(__m128d)(A), (int)(R))
5974
+ ((int)__builtin_ia32_vcvtsd2si32((__v2df)(__m128d)(A), (int)(R)))
5971
5975
 
5972
5976
  #define _mm_cvt_roundsd_i32(A, R) \
5973
- (int)__builtin_ia32_vcvtsd2si32((__v2df)(__m128d)(A), (int)(R))
5977
+ ((int)__builtin_ia32_vcvtsd2si32((__v2df)(__m128d)(A), (int)(R)))
5974
5978
 
5975
5979
  #define _mm_cvt_roundsd_u32(A, R) \
5976
- (unsigned int)__builtin_ia32_vcvtsd2usi32((__v2df)(__m128d)(A), (int)(R))
5980
+ ((unsigned int)__builtin_ia32_vcvtsd2usi32((__v2df)(__m128d)(A), (int)(R)))
5977
5981
 
5978
5982
  static __inline__ unsigned __DEFAULT_FN_ATTRS128
5979
5983
  _mm_cvtsd_u32 (__m128d __A)
@@ -5984,8 +5988,8 @@ _mm_cvtsd_u32 (__m128d __A)
5984
5988
 
5985
5989
  #ifdef __x86_64__
5986
5990
  #define _mm_cvt_roundsd_u64(A, R) \
5987
- (unsigned long long)__builtin_ia32_vcvtsd2usi64((__v2df)(__m128d)(A), \
5988
- (int)(R))
5991
+ ((unsigned long long)__builtin_ia32_vcvtsd2usi64((__v2df)(__m128d)(A), \
5992
+ (int)(R)))
5989
5993
 
5990
5994
  static __inline__ unsigned long long __DEFAULT_FN_ATTRS128
5991
5995
  _mm_cvtsd_u64 (__m128d __A)
@@ -5997,21 +6001,21 @@ _mm_cvtsd_u64 (__m128d __A)
5997
6001
  #endif
5998
6002
 
5999
6003
  #define _mm_cvt_roundss_si32(A, R) \
6000
- (int)__builtin_ia32_vcvtss2si32((__v4sf)(__m128)(A), (int)(R))
6004
+ ((int)__builtin_ia32_vcvtss2si32((__v4sf)(__m128)(A), (int)(R)))
6001
6005
 
6002
6006
  #define _mm_cvt_roundss_i32(A, R) \
6003
- (int)__builtin_ia32_vcvtss2si32((__v4sf)(__m128)(A), (int)(R))
6007
+ ((int)__builtin_ia32_vcvtss2si32((__v4sf)(__m128)(A), (int)(R)))
6004
6008
 
6005
6009
  #ifdef __x86_64__
6006
6010
  #define _mm_cvt_roundss_si64(A, R) \
6007
- (long long)__builtin_ia32_vcvtss2si64((__v4sf)(__m128)(A), (int)(R))
6011
+ ((long long)__builtin_ia32_vcvtss2si64((__v4sf)(__m128)(A), (int)(R)))
6008
6012
 
6009
6013
  #define _mm_cvt_roundss_i64(A, R) \
6010
- (long long)__builtin_ia32_vcvtss2si64((__v4sf)(__m128)(A), (int)(R))
6014
+ ((long long)__builtin_ia32_vcvtss2si64((__v4sf)(__m128)(A), (int)(R)))
6011
6015
  #endif
6012
6016
 
6013
6017
  #define _mm_cvt_roundss_u32(A, R) \
6014
- (unsigned int)__builtin_ia32_vcvtss2usi32((__v4sf)(__m128)(A), (int)(R))
6018
+ ((unsigned int)__builtin_ia32_vcvtss2usi32((__v4sf)(__m128)(A), (int)(R)))
6015
6019
 
6016
6020
  static __inline__ unsigned __DEFAULT_FN_ATTRS128
6017
6021
  _mm_cvtss_u32 (__m128 __A)
@@ -6022,8 +6026,8 @@ _mm_cvtss_u32 (__m128 __A)
6022
6026
 
6023
6027
  #ifdef __x86_64__
6024
6028
  #define _mm_cvt_roundss_u64(A, R) \
6025
- (unsigned long long)__builtin_ia32_vcvtss2usi64((__v4sf)(__m128)(A), \
6026
- (int)(R))
6029
+ ((unsigned long long)__builtin_ia32_vcvtss2usi64((__v4sf)(__m128)(A), \
6030
+ (int)(R)))
6027
6031
 
6028
6032
  static __inline__ unsigned long long __DEFAULT_FN_ATTRS128
6029
6033
  _mm_cvtss_u64 (__m128 __A)
@@ -6035,10 +6039,10 @@ _mm_cvtss_u64 (__m128 __A)
6035
6039
  #endif
6036
6040
 
6037
6041
  #define _mm_cvtt_roundsd_i32(A, R) \
6038
- (int)__builtin_ia32_vcvttsd2si32((__v2df)(__m128d)(A), (int)(R))
6042
+ ((int)__builtin_ia32_vcvttsd2si32((__v2df)(__m128d)(A), (int)(R)))
6039
6043
 
6040
6044
  #define _mm_cvtt_roundsd_si32(A, R) \
6041
- (int)__builtin_ia32_vcvttsd2si32((__v2df)(__m128d)(A), (int)(R))
6045
+ ((int)__builtin_ia32_vcvttsd2si32((__v2df)(__m128d)(A), (int)(R)))
6042
6046
 
6043
6047
  static __inline__ int __DEFAULT_FN_ATTRS128
6044
6048
  _mm_cvttsd_i32 (__m128d __A)
@@ -6049,10 +6053,10 @@ _mm_cvttsd_i32 (__m128d __A)
6049
6053
 
6050
6054
  #ifdef __x86_64__
6051
6055
  #define _mm_cvtt_roundsd_si64(A, R) \
6052
- (long long)__builtin_ia32_vcvttsd2si64((__v2df)(__m128d)(A), (int)(R))
6056
+ ((long long)__builtin_ia32_vcvttsd2si64((__v2df)(__m128d)(A), (int)(R)))
6053
6057
 
6054
6058
  #define _mm_cvtt_roundsd_i64(A, R) \
6055
- (long long)__builtin_ia32_vcvttsd2si64((__v2df)(__m128d)(A), (int)(R))
6059
+ ((long long)__builtin_ia32_vcvttsd2si64((__v2df)(__m128d)(A), (int)(R)))
6056
6060
 
6057
6061
  static __inline__ long long __DEFAULT_FN_ATTRS128
6058
6062
  _mm_cvttsd_i64 (__m128d __A)
@@ -6063,7 +6067,7 @@ _mm_cvttsd_i64 (__m128d __A)
6063
6067
  #endif
6064
6068
 
6065
6069
  #define _mm_cvtt_roundsd_u32(A, R) \
6066
- (unsigned int)__builtin_ia32_vcvttsd2usi32((__v2df)(__m128d)(A), (int)(R))
6070
+ ((unsigned int)__builtin_ia32_vcvttsd2usi32((__v2df)(__m128d)(A), (int)(R)))
6067
6071
 
6068
6072
  static __inline__ unsigned __DEFAULT_FN_ATTRS128
6069
6073
  _mm_cvttsd_u32 (__m128d __A)
@@ -6074,8 +6078,8 @@ _mm_cvttsd_u32 (__m128d __A)
6074
6078
 
6075
6079
  #ifdef __x86_64__
6076
6080
  #define _mm_cvtt_roundsd_u64(A, R) \
6077
- (unsigned long long)__builtin_ia32_vcvttsd2usi64((__v2df)(__m128d)(A), \
6078
- (int)(R))
6081
+ ((unsigned long long)__builtin_ia32_vcvttsd2usi64((__v2df)(__m128d)(A), \
6082
+ (int)(R)))
6079
6083
 
6080
6084
  static __inline__ unsigned long long __DEFAULT_FN_ATTRS128
6081
6085
  _mm_cvttsd_u64 (__m128d __A)
@@ -6087,10 +6091,10 @@ _mm_cvttsd_u64 (__m128d __A)
6087
6091
  #endif
6088
6092
 
6089
6093
  #define _mm_cvtt_roundss_i32(A, R) \
6090
- (int)__builtin_ia32_vcvttss2si32((__v4sf)(__m128)(A), (int)(R))
6094
+ ((int)__builtin_ia32_vcvttss2si32((__v4sf)(__m128)(A), (int)(R)))
6091
6095
 
6092
6096
  #define _mm_cvtt_roundss_si32(A, R) \
6093
- (int)__builtin_ia32_vcvttss2si32((__v4sf)(__m128)(A), (int)(R))
6097
+ ((int)__builtin_ia32_vcvttss2si32((__v4sf)(__m128)(A), (int)(R)))
6094
6098
 
6095
6099
  static __inline__ int __DEFAULT_FN_ATTRS128
6096
6100
  _mm_cvttss_i32 (__m128 __A)
@@ -6101,10 +6105,10 @@ _mm_cvttss_i32 (__m128 __A)
6101
6105
 
6102
6106
  #ifdef __x86_64__
6103
6107
  #define _mm_cvtt_roundss_i64(A, R) \
6104
- (long long)__builtin_ia32_vcvttss2si64((__v4sf)(__m128)(A), (int)(R))
6108
+ ((long long)__builtin_ia32_vcvttss2si64((__v4sf)(__m128)(A), (int)(R)))
6105
6109
 
6106
6110
  #define _mm_cvtt_roundss_si64(A, R) \
6107
- (long long)__builtin_ia32_vcvttss2si64((__v4sf)(__m128)(A), (int)(R))
6111
+ ((long long)__builtin_ia32_vcvttss2si64((__v4sf)(__m128)(A), (int)(R)))
6108
6112
 
6109
6113
  static __inline__ long long __DEFAULT_FN_ATTRS128
6110
6114
  _mm_cvttss_i64 (__m128 __A)
@@ -6115,7 +6119,7 @@ _mm_cvttss_i64 (__m128 __A)
6115
6119
  #endif
6116
6120
 
6117
6121
  #define _mm_cvtt_roundss_u32(A, R) \
6118
- (unsigned int)__builtin_ia32_vcvttss2usi32((__v4sf)(__m128)(A), (int)(R))
6122
+ ((unsigned int)__builtin_ia32_vcvttss2usi32((__v4sf)(__m128)(A), (int)(R)))
6119
6123
 
6120
6124
  static __inline__ unsigned __DEFAULT_FN_ATTRS128
6121
6125
  _mm_cvttss_u32 (__m128 __A)
@@ -6126,8 +6130,8 @@ _mm_cvttss_u32 (__m128 __A)
6126
6130
 
6127
6131
  #ifdef __x86_64__
6128
6132
  #define _mm_cvtt_roundss_u64(A, R) \
6129
- (unsigned long long)__builtin_ia32_vcvttss2usi64((__v4sf)(__m128)(A), \
6130
- (int)(R))
6133
+ ((unsigned long long)__builtin_ia32_vcvttss2usi64((__v4sf)(__m128)(A), \
6134
+ (int)(R)))
6131
6135
 
6132
6136
  static __inline__ unsigned long long __DEFAULT_FN_ATTRS128
6133
6137
  _mm_cvttss_u64 (__m128 __A)
@@ -6139,30 +6143,30 @@ _mm_cvttss_u64 (__m128 __A)
6139
6143
  #endif
6140
6144
 
6141
6145
  #define _mm512_permute_pd(X, C) \
6142
- (__m512d)__builtin_ia32_vpermilpd512((__v8df)(__m512d)(X), (int)(C))
6146
+ ((__m512d)__builtin_ia32_vpermilpd512((__v8df)(__m512d)(X), (int)(C)))
6143
6147
 
6144
6148
  #define _mm512_mask_permute_pd(W, U, X, C) \
6145
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6146
- (__v8df)_mm512_permute_pd((X), (C)), \
6147
- (__v8df)(__m512d)(W))
6149
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6150
+ (__v8df)_mm512_permute_pd((X), (C)), \
6151
+ (__v8df)(__m512d)(W)))
6148
6152
 
6149
6153
  #define _mm512_maskz_permute_pd(U, X, C) \
6150
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6151
- (__v8df)_mm512_permute_pd((X), (C)), \
6152
- (__v8df)_mm512_setzero_pd())
6154
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6155
+ (__v8df)_mm512_permute_pd((X), (C)), \
6156
+ (__v8df)_mm512_setzero_pd()))
6153
6157
 
6154
6158
  #define _mm512_permute_ps(X, C) \
6155
- (__m512)__builtin_ia32_vpermilps512((__v16sf)(__m512)(X), (int)(C))
6159
+ ((__m512)__builtin_ia32_vpermilps512((__v16sf)(__m512)(X), (int)(C)))
6156
6160
 
6157
6161
  #define _mm512_mask_permute_ps(W, U, X, C) \
6158
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6159
- (__v16sf)_mm512_permute_ps((X), (C)), \
6160
- (__v16sf)(__m512)(W))
6162
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6163
+ (__v16sf)_mm512_permute_ps((X), (C)), \
6164
+ (__v16sf)(__m512)(W)))
6161
6165
 
6162
6166
  #define _mm512_maskz_permute_ps(U, X, C) \
6163
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6164
- (__v16sf)_mm512_permute_ps((X), (C)), \
6165
- (__v16sf)_mm512_setzero_ps())
6167
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6168
+ (__v16sf)_mm512_permute_ps((X), (C)), \
6169
+ (__v16sf)_mm512_setzero_ps()))
6166
6170
 
6167
6171
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
6168
6172
  _mm512_permutevar_pd(__m512d __A, __m512i __C)
@@ -6274,19 +6278,19 @@ _mm512_maskz_permutex2var_ps(__mmask16 __U, __m512 __A, __m512i __I, __m512 __B)
6274
6278
 
6275
6279
 
6276
6280
  #define _mm512_cvtt_roundpd_epu32(A, R) \
6277
- (__m256i)__builtin_ia32_cvttpd2udq512_mask((__v8df)(__m512d)(A), \
6278
- (__v8si)_mm256_undefined_si256(), \
6279
- (__mmask8)-1, (int)(R))
6281
+ ((__m256i)__builtin_ia32_cvttpd2udq512_mask((__v8df)(__m512d)(A), \
6282
+ (__v8si)_mm256_undefined_si256(), \
6283
+ (__mmask8)-1, (int)(R)))
6280
6284
 
6281
6285
  #define _mm512_mask_cvtt_roundpd_epu32(W, U, A, R) \
6282
- (__m256i)__builtin_ia32_cvttpd2udq512_mask((__v8df)(__m512d)(A), \
6283
- (__v8si)(__m256i)(W), \
6284
- (__mmask8)(U), (int)(R))
6286
+ ((__m256i)__builtin_ia32_cvttpd2udq512_mask((__v8df)(__m512d)(A), \
6287
+ (__v8si)(__m256i)(W), \
6288
+ (__mmask8)(U), (int)(R)))
6285
6289
 
6286
6290
  #define _mm512_maskz_cvtt_roundpd_epu32(U, A, R) \
6287
- (__m256i)__builtin_ia32_cvttpd2udq512_mask((__v8df)(__m512d)(A), \
6288
- (__v8si)_mm256_setzero_si256(), \
6289
- (__mmask8)(U), (int)(R))
6291
+ ((__m256i)__builtin_ia32_cvttpd2udq512_mask((__v8df)(__m512d)(A), \
6292
+ (__v8si)_mm256_setzero_si256(), \
6293
+ (__mmask8)(U), (int)(R)))
6290
6294
 
6291
6295
  static __inline__ __m256i __DEFAULT_FN_ATTRS512
6292
6296
  _mm512_cvttpd_epu32 (__m512d __A)
@@ -6318,106 +6322,106 @@ _mm512_maskz_cvttpd_epu32 (__mmask8 __U, __m512d __A)
6318
6322
  }
6319
6323
 
6320
6324
  #define _mm_roundscale_round_sd(A, B, imm, R) \
6321
- (__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6322
- (__v2df)(__m128d)(B), \
6323
- (__v2df)_mm_setzero_pd(), \
6324
- (__mmask8)-1, (int)(imm), \
6325
- (int)(R))
6325
+ ((__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6326
+ (__v2df)(__m128d)(B), \
6327
+ (__v2df)_mm_setzero_pd(), \
6328
+ (__mmask8)-1, (int)(imm), \
6329
+ (int)(R)))
6326
6330
 
6327
6331
  #define _mm_roundscale_sd(A, B, imm) \
6328
- (__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6329
- (__v2df)(__m128d)(B), \
6330
- (__v2df)_mm_setzero_pd(), \
6331
- (__mmask8)-1, (int)(imm), \
6332
- _MM_FROUND_CUR_DIRECTION)
6332
+ ((__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6333
+ (__v2df)(__m128d)(B), \
6334
+ (__v2df)_mm_setzero_pd(), \
6335
+ (__mmask8)-1, (int)(imm), \
6336
+ _MM_FROUND_CUR_DIRECTION))
6333
6337
 
6334
6338
  #define _mm_mask_roundscale_sd(W, U, A, B, imm) \
6335
- (__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6336
- (__v2df)(__m128d)(B), \
6337
- (__v2df)(__m128d)(W), \
6338
- (__mmask8)(U), (int)(imm), \
6339
- _MM_FROUND_CUR_DIRECTION)
6339
+ ((__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6340
+ (__v2df)(__m128d)(B), \
6341
+ (__v2df)(__m128d)(W), \
6342
+ (__mmask8)(U), (int)(imm), \
6343
+ _MM_FROUND_CUR_DIRECTION))
6340
6344
 
6341
6345
  #define _mm_mask_roundscale_round_sd(W, U, A, B, I, R) \
6342
- (__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6343
- (__v2df)(__m128d)(B), \
6344
- (__v2df)(__m128d)(W), \
6345
- (__mmask8)(U), (int)(I), \
6346
- (int)(R))
6346
+ ((__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6347
+ (__v2df)(__m128d)(B), \
6348
+ (__v2df)(__m128d)(W), \
6349
+ (__mmask8)(U), (int)(I), \
6350
+ (int)(R)))
6347
6351
 
6348
6352
  #define _mm_maskz_roundscale_sd(U, A, B, I) \
6349
- (__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6350
- (__v2df)(__m128d)(B), \
6351
- (__v2df)_mm_setzero_pd(), \
6352
- (__mmask8)(U), (int)(I), \
6353
- _MM_FROUND_CUR_DIRECTION)
6353
+ ((__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6354
+ (__v2df)(__m128d)(B), \
6355
+ (__v2df)_mm_setzero_pd(), \
6356
+ (__mmask8)(U), (int)(I), \
6357
+ _MM_FROUND_CUR_DIRECTION))
6354
6358
 
6355
6359
  #define _mm_maskz_roundscale_round_sd(U, A, B, I, R) \
6356
- (__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6357
- (__v2df)(__m128d)(B), \
6358
- (__v2df)_mm_setzero_pd(), \
6359
- (__mmask8)(U), (int)(I), \
6360
- (int)(R))
6360
+ ((__m128d)__builtin_ia32_rndscalesd_round_mask((__v2df)(__m128d)(A), \
6361
+ (__v2df)(__m128d)(B), \
6362
+ (__v2df)_mm_setzero_pd(), \
6363
+ (__mmask8)(U), (int)(I), \
6364
+ (int)(R)))
6361
6365
 
6362
6366
  #define _mm_roundscale_round_ss(A, B, imm, R) \
6363
- (__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6364
- (__v4sf)(__m128)(B), \
6365
- (__v4sf)_mm_setzero_ps(), \
6366
- (__mmask8)-1, (int)(imm), \
6367
- (int)(R))
6367
+ ((__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6368
+ (__v4sf)(__m128)(B), \
6369
+ (__v4sf)_mm_setzero_ps(), \
6370
+ (__mmask8)-1, (int)(imm), \
6371
+ (int)(R)))
6368
6372
 
6369
6373
  #define _mm_roundscale_ss(A, B, imm) \
6370
- (__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6371
- (__v4sf)(__m128)(B), \
6372
- (__v4sf)_mm_setzero_ps(), \
6373
- (__mmask8)-1, (int)(imm), \
6374
- _MM_FROUND_CUR_DIRECTION)
6374
+ ((__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6375
+ (__v4sf)(__m128)(B), \
6376
+ (__v4sf)_mm_setzero_ps(), \
6377
+ (__mmask8)-1, (int)(imm), \
6378
+ _MM_FROUND_CUR_DIRECTION))
6375
6379
 
6376
6380
  #define _mm_mask_roundscale_ss(W, U, A, B, I) \
6377
- (__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6378
- (__v4sf)(__m128)(B), \
6379
- (__v4sf)(__m128)(W), \
6380
- (__mmask8)(U), (int)(I), \
6381
- _MM_FROUND_CUR_DIRECTION)
6381
+ ((__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6382
+ (__v4sf)(__m128)(B), \
6383
+ (__v4sf)(__m128)(W), \
6384
+ (__mmask8)(U), (int)(I), \
6385
+ _MM_FROUND_CUR_DIRECTION))
6382
6386
 
6383
6387
  #define _mm_mask_roundscale_round_ss(W, U, A, B, I, R) \
6384
- (__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6385
- (__v4sf)(__m128)(B), \
6386
- (__v4sf)(__m128)(W), \
6387
- (__mmask8)(U), (int)(I), \
6388
- (int)(R))
6388
+ ((__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6389
+ (__v4sf)(__m128)(B), \
6390
+ (__v4sf)(__m128)(W), \
6391
+ (__mmask8)(U), (int)(I), \
6392
+ (int)(R)))
6389
6393
 
6390
6394
  #define _mm_maskz_roundscale_ss(U, A, B, I) \
6391
- (__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6392
- (__v4sf)(__m128)(B), \
6393
- (__v4sf)_mm_setzero_ps(), \
6394
- (__mmask8)(U), (int)(I), \
6395
- _MM_FROUND_CUR_DIRECTION)
6395
+ ((__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6396
+ (__v4sf)(__m128)(B), \
6397
+ (__v4sf)_mm_setzero_ps(), \
6398
+ (__mmask8)(U), (int)(I), \
6399
+ _MM_FROUND_CUR_DIRECTION))
6396
6400
 
6397
6401
  #define _mm_maskz_roundscale_round_ss(U, A, B, I, R) \
6398
- (__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6399
- (__v4sf)(__m128)(B), \
6400
- (__v4sf)_mm_setzero_ps(), \
6401
- (__mmask8)(U), (int)(I), \
6402
- (int)(R))
6402
+ ((__m128)__builtin_ia32_rndscaless_round_mask((__v4sf)(__m128)(A), \
6403
+ (__v4sf)(__m128)(B), \
6404
+ (__v4sf)_mm_setzero_ps(), \
6405
+ (__mmask8)(U), (int)(I), \
6406
+ (int)(R)))
6403
6407
 
6404
6408
  #define _mm512_scalef_round_pd(A, B, R) \
6405
- (__m512d)__builtin_ia32_scalefpd512_mask((__v8df)(__m512d)(A), \
6406
- (__v8df)(__m512d)(B), \
6407
- (__v8df)_mm512_undefined_pd(), \
6408
- (__mmask8)-1, (int)(R))
6409
+ ((__m512d)__builtin_ia32_scalefpd512_mask((__v8df)(__m512d)(A), \
6410
+ (__v8df)(__m512d)(B), \
6411
+ (__v8df)_mm512_undefined_pd(), \
6412
+ (__mmask8)-1, (int)(R)))
6409
6413
 
6410
6414
  #define _mm512_mask_scalef_round_pd(W, U, A, B, R) \
6411
- (__m512d)__builtin_ia32_scalefpd512_mask((__v8df)(__m512d)(A), \
6412
- (__v8df)(__m512d)(B), \
6413
- (__v8df)(__m512d)(W), \
6414
- (__mmask8)(U), (int)(R))
6415
+ ((__m512d)__builtin_ia32_scalefpd512_mask((__v8df)(__m512d)(A), \
6416
+ (__v8df)(__m512d)(B), \
6417
+ (__v8df)(__m512d)(W), \
6418
+ (__mmask8)(U), (int)(R)))
6415
6419
 
6416
6420
  #define _mm512_maskz_scalef_round_pd(U, A, B, R) \
6417
- (__m512d)__builtin_ia32_scalefpd512_mask((__v8df)(__m512d)(A), \
6418
- (__v8df)(__m512d)(B), \
6419
- (__v8df)_mm512_setzero_pd(), \
6420
- (__mmask8)(U), (int)(R))
6421
+ ((__m512d)__builtin_ia32_scalefpd512_mask((__v8df)(__m512d)(A), \
6422
+ (__v8df)(__m512d)(B), \
6423
+ (__v8df)_mm512_setzero_pd(), \
6424
+ (__mmask8)(U), (int)(R)))
6421
6425
 
6422
6426
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
6423
6427
  _mm512_scalef_pd (__m512d __A, __m512d __B)
@@ -6452,22 +6456,22 @@ _mm512_maskz_scalef_pd (__mmask8 __U, __m512d __A, __m512d __B)
6452
6456
  }
6453
6457
 
6454
6458
  #define _mm512_scalef_round_ps(A, B, R) \
6455
- (__m512)__builtin_ia32_scalefps512_mask((__v16sf)(__m512)(A), \
6456
- (__v16sf)(__m512)(B), \
6457
- (__v16sf)_mm512_undefined_ps(), \
6458
- (__mmask16)-1, (int)(R))
6459
+ ((__m512)__builtin_ia32_scalefps512_mask((__v16sf)(__m512)(A), \
6460
+ (__v16sf)(__m512)(B), \
6461
+ (__v16sf)_mm512_undefined_ps(), \
6462
+ (__mmask16)-1, (int)(R)))
6459
6463
 
6460
6464
  #define _mm512_mask_scalef_round_ps(W, U, A, B, R) \
6461
- (__m512)__builtin_ia32_scalefps512_mask((__v16sf)(__m512)(A), \
6462
- (__v16sf)(__m512)(B), \
6463
- (__v16sf)(__m512)(W), \
6464
- (__mmask16)(U), (int)(R))
6465
+ ((__m512)__builtin_ia32_scalefps512_mask((__v16sf)(__m512)(A), \
6466
+ (__v16sf)(__m512)(B), \
6467
+ (__v16sf)(__m512)(W), \
6468
+ (__mmask16)(U), (int)(R)))
6465
6469
 
6466
6470
  #define _mm512_maskz_scalef_round_ps(U, A, B, R) \
6467
- (__m512)__builtin_ia32_scalefps512_mask((__v16sf)(__m512)(A), \
6468
- (__v16sf)(__m512)(B), \
6469
- (__v16sf)_mm512_setzero_ps(), \
6470
- (__mmask16)(U), (int)(R))
6471
+ ((__m512)__builtin_ia32_scalefps512_mask((__v16sf)(__m512)(A), \
6472
+ (__v16sf)(__m512)(B), \
6473
+ (__v16sf)_mm512_setzero_ps(), \
6474
+ (__mmask16)(U), (int)(R)))
6471
6475
 
6472
6476
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
6473
6477
  _mm512_scalef_ps (__m512 __A, __m512 __B)
@@ -6502,10 +6506,10 @@ _mm512_maskz_scalef_ps (__mmask16 __U, __m512 __A, __m512 __B)
6502
6506
  }
6503
6507
 
6504
6508
  #define _mm_scalef_round_sd(A, B, R) \
6505
- (__m128d)__builtin_ia32_scalefsd_round_mask((__v2df)(__m128d)(A), \
6506
- (__v2df)(__m128d)(B), \
6507
- (__v2df)_mm_setzero_pd(), \
6508
- (__mmask8)-1, (int)(R))
6509
+ ((__m128d)__builtin_ia32_scalefsd_round_mask((__v2df)(__m128d)(A), \
6510
+ (__v2df)(__m128d)(B), \
6511
+ (__v2df)_mm_setzero_pd(), \
6512
+ (__mmask8)-1, (int)(R)))
6509
6513
 
6510
6514
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
6511
6515
  _mm_scalef_sd (__m128d __A, __m128d __B)
@@ -6527,10 +6531,10 @@ _mm_mask_scalef_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
6527
6531
  }
6528
6532
 
6529
6533
  #define _mm_mask_scalef_round_sd(W, U, A, B, R) \
6530
- (__m128d)__builtin_ia32_scalefsd_round_mask((__v2df)(__m128d)(A), \
6531
- (__v2df)(__m128d)(B), \
6532
- (__v2df)(__m128d)(W), \
6533
- (__mmask8)(U), (int)(R))
6534
+ ((__m128d)__builtin_ia32_scalefsd_round_mask((__v2df)(__m128d)(A), \
6535
+ (__v2df)(__m128d)(B), \
6536
+ (__v2df)(__m128d)(W), \
6537
+ (__mmask8)(U), (int)(R)))
6534
6538
 
6535
6539
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
6536
6540
  _mm_maskz_scalef_sd (__mmask8 __U, __m128d __A, __m128d __B)
@@ -6543,16 +6547,16 @@ _mm_maskz_scalef_sd (__mmask8 __U, __m128d __A, __m128d __B)
6543
6547
  }
6544
6548
 
6545
6549
  #define _mm_maskz_scalef_round_sd(U, A, B, R) \
6546
- (__m128d)__builtin_ia32_scalefsd_round_mask((__v2df)(__m128d)(A), \
6547
- (__v2df)(__m128d)(B), \
6548
- (__v2df)_mm_setzero_pd(), \
6549
- (__mmask8)(U), (int)(R))
6550
+ ((__m128d)__builtin_ia32_scalefsd_round_mask((__v2df)(__m128d)(A), \
6551
+ (__v2df)(__m128d)(B), \
6552
+ (__v2df)_mm_setzero_pd(), \
6553
+ (__mmask8)(U), (int)(R)))
6550
6554
 
6551
6555
  #define _mm_scalef_round_ss(A, B, R) \
6552
- (__m128)__builtin_ia32_scalefss_round_mask((__v4sf)(__m128)(A), \
6553
- (__v4sf)(__m128)(B), \
6554
- (__v4sf)_mm_setzero_ps(), \
6555
- (__mmask8)-1, (int)(R))
6556
+ ((__m128)__builtin_ia32_scalefss_round_mask((__v4sf)(__m128)(A), \
6557
+ (__v4sf)(__m128)(B), \
6558
+ (__v4sf)_mm_setzero_ps(), \
6559
+ (__mmask8)-1, (int)(R)))
6556
6560
 
6557
6561
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
6558
6562
  _mm_scalef_ss (__m128 __A, __m128 __B)
@@ -6574,10 +6578,10 @@ _mm_mask_scalef_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
6574
6578
  }
6575
6579
 
6576
6580
  #define _mm_mask_scalef_round_ss(W, U, A, B, R) \
6577
- (__m128)__builtin_ia32_scalefss_round_mask((__v4sf)(__m128)(A), \
6578
- (__v4sf)(__m128)(B), \
6579
- (__v4sf)(__m128)(W), \
6580
- (__mmask8)(U), (int)(R))
6581
+ ((__m128)__builtin_ia32_scalefss_round_mask((__v4sf)(__m128)(A), \
6582
+ (__v4sf)(__m128)(B), \
6583
+ (__v4sf)(__m128)(W), \
6584
+ (__mmask8)(U), (int)(R)))
6581
6585
 
6582
6586
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
6583
6587
  _mm_maskz_scalef_ss (__mmask8 __U, __m128 __A, __m128 __B)
@@ -6590,11 +6594,11 @@ _mm_maskz_scalef_ss (__mmask8 __U, __m128 __A, __m128 __B)
6590
6594
  }
6591
6595
 
6592
6596
  #define _mm_maskz_scalef_round_ss(U, A, B, R) \
6593
- (__m128)__builtin_ia32_scalefss_round_mask((__v4sf)(__m128)(A), \
6594
- (__v4sf)(__m128)(B), \
6595
- (__v4sf)_mm_setzero_ps(), \
6596
- (__mmask8)(U), \
6597
- (int)(R))
6597
+ ((__m128)__builtin_ia32_scalefss_round_mask((__v4sf)(__m128)(A), \
6598
+ (__v4sf)(__m128)(B), \
6599
+ (__v4sf)_mm_setzero_ps(), \
6600
+ (__mmask8)(U), \
6601
+ (int)(R)))
6598
6602
 
6599
6603
  static __inline__ __m512i __DEFAULT_FN_ATTRS512
6600
6604
  _mm512_srai_epi32(__m512i __A, unsigned int __B)
@@ -6642,94 +6646,94 @@ _mm512_maskz_srai_epi64(__mmask8 __U, __m512i __A, unsigned int __B)
6642
6646
  }
6643
6647
 
6644
6648
  #define _mm512_shuffle_f32x4(A, B, imm) \
6645
- (__m512)__builtin_ia32_shuf_f32x4((__v16sf)(__m512)(A), \
6646
- (__v16sf)(__m512)(B), (int)(imm))
6649
+ ((__m512)__builtin_ia32_shuf_f32x4((__v16sf)(__m512)(A), \
6650
+ (__v16sf)(__m512)(B), (int)(imm)))
6647
6651
 
6648
6652
  #define _mm512_mask_shuffle_f32x4(W, U, A, B, imm) \
6649
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6650
- (__v16sf)_mm512_shuffle_f32x4((A), (B), (imm)), \
6651
- (__v16sf)(__m512)(W))
6653
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6654
+ (__v16sf)_mm512_shuffle_f32x4((A), (B), (imm)), \
6655
+ (__v16sf)(__m512)(W)))
6652
6656
 
6653
6657
  #define _mm512_maskz_shuffle_f32x4(U, A, B, imm) \
6654
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6655
- (__v16sf)_mm512_shuffle_f32x4((A), (B), (imm)), \
6656
- (__v16sf)_mm512_setzero_ps())
6658
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6659
+ (__v16sf)_mm512_shuffle_f32x4((A), (B), (imm)), \
6660
+ (__v16sf)_mm512_setzero_ps()))
6657
6661
 
6658
6662
  #define _mm512_shuffle_f64x2(A, B, imm) \
6659
- (__m512d)__builtin_ia32_shuf_f64x2((__v8df)(__m512d)(A), \
6660
- (__v8df)(__m512d)(B), (int)(imm))
6663
+ ((__m512d)__builtin_ia32_shuf_f64x2((__v8df)(__m512d)(A), \
6664
+ (__v8df)(__m512d)(B), (int)(imm)))
6661
6665
 
6662
6666
  #define _mm512_mask_shuffle_f64x2(W, U, A, B, imm) \
6663
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6664
- (__v8df)_mm512_shuffle_f64x2((A), (B), (imm)), \
6665
- (__v8df)(__m512d)(W))
6667
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6668
+ (__v8df)_mm512_shuffle_f64x2((A), (B), (imm)), \
6669
+ (__v8df)(__m512d)(W)))
6666
6670
 
6667
6671
  #define _mm512_maskz_shuffle_f64x2(U, A, B, imm) \
6668
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6669
- (__v8df)_mm512_shuffle_f64x2((A), (B), (imm)), \
6670
- (__v8df)_mm512_setzero_pd())
6672
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6673
+ (__v8df)_mm512_shuffle_f64x2((A), (B), (imm)), \
6674
+ (__v8df)_mm512_setzero_pd()))
6671
6675
 
6672
6676
  #define _mm512_shuffle_i32x4(A, B, imm) \
6673
- (__m512i)__builtin_ia32_shuf_i32x4((__v16si)(__m512i)(A), \
6674
- (__v16si)(__m512i)(B), (int)(imm))
6677
+ ((__m512i)__builtin_ia32_shuf_i32x4((__v16si)(__m512i)(A), \
6678
+ (__v16si)(__m512i)(B), (int)(imm)))
6675
6679
 
6676
6680
  #define _mm512_mask_shuffle_i32x4(W, U, A, B, imm) \
6677
- (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
6678
- (__v16si)_mm512_shuffle_i32x4((A), (B), (imm)), \
6679
- (__v16si)(__m512i)(W))
6681
+ ((__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
6682
+ (__v16si)_mm512_shuffle_i32x4((A), (B), (imm)), \
6683
+ (__v16si)(__m512i)(W)))
6680
6684
 
6681
6685
  #define _mm512_maskz_shuffle_i32x4(U, A, B, imm) \
6682
- (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
6683
- (__v16si)_mm512_shuffle_i32x4((A), (B), (imm)), \
6684
- (__v16si)_mm512_setzero_si512())
6686
+ ((__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
6687
+ (__v16si)_mm512_shuffle_i32x4((A), (B), (imm)), \
6688
+ (__v16si)_mm512_setzero_si512()))
6685
6689
 
6686
6690
  #define _mm512_shuffle_i64x2(A, B, imm) \
6687
- (__m512i)__builtin_ia32_shuf_i64x2((__v8di)(__m512i)(A), \
6688
- (__v8di)(__m512i)(B), (int)(imm))
6691
+ ((__m512i)__builtin_ia32_shuf_i64x2((__v8di)(__m512i)(A), \
6692
+ (__v8di)(__m512i)(B), (int)(imm)))
6689
6693
 
6690
6694
  #define _mm512_mask_shuffle_i64x2(W, U, A, B, imm) \
6691
- (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
6692
- (__v8di)_mm512_shuffle_i64x2((A), (B), (imm)), \
6693
- (__v8di)(__m512i)(W))
6695
+ ((__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
6696
+ (__v8di)_mm512_shuffle_i64x2((A), (B), (imm)), \
6697
+ (__v8di)(__m512i)(W)))
6694
6698
 
6695
6699
  #define _mm512_maskz_shuffle_i64x2(U, A, B, imm) \
6696
- (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
6697
- (__v8di)_mm512_shuffle_i64x2((A), (B), (imm)), \
6698
- (__v8di)_mm512_setzero_si512())
6700
+ ((__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
6701
+ (__v8di)_mm512_shuffle_i64x2((A), (B), (imm)), \
6702
+ (__v8di)_mm512_setzero_si512()))
6699
6703
 
6700
6704
  #define _mm512_shuffle_pd(A, B, M) \
6701
- (__m512d)__builtin_ia32_shufpd512((__v8df)(__m512d)(A), \
6702
- (__v8df)(__m512d)(B), (int)(M))
6705
+ ((__m512d)__builtin_ia32_shufpd512((__v8df)(__m512d)(A), \
6706
+ (__v8df)(__m512d)(B), (int)(M)))
6703
6707
 
6704
6708
  #define _mm512_mask_shuffle_pd(W, U, A, B, M) \
6705
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6706
- (__v8df)_mm512_shuffle_pd((A), (B), (M)), \
6707
- (__v8df)(__m512d)(W))
6709
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6710
+ (__v8df)_mm512_shuffle_pd((A), (B), (M)), \
6711
+ (__v8df)(__m512d)(W)))
6708
6712
 
6709
6713
  #define _mm512_maskz_shuffle_pd(U, A, B, M) \
6710
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6711
- (__v8df)_mm512_shuffle_pd((A), (B), (M)), \
6712
- (__v8df)_mm512_setzero_pd())
6714
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
6715
+ (__v8df)_mm512_shuffle_pd((A), (B), (M)), \
6716
+ (__v8df)_mm512_setzero_pd()))
6713
6717
 
6714
6718
  #define _mm512_shuffle_ps(A, B, M) \
6715
- (__m512)__builtin_ia32_shufps512((__v16sf)(__m512)(A), \
6716
- (__v16sf)(__m512)(B), (int)(M))
6719
+ ((__m512)__builtin_ia32_shufps512((__v16sf)(__m512)(A), \
6720
+ (__v16sf)(__m512)(B), (int)(M)))
6717
6721
 
6718
6722
  #define _mm512_mask_shuffle_ps(W, U, A, B, M) \
6719
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6720
- (__v16sf)_mm512_shuffle_ps((A), (B), (M)), \
6721
- (__v16sf)(__m512)(W))
6723
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6724
+ (__v16sf)_mm512_shuffle_ps((A), (B), (M)), \
6725
+ (__v16sf)(__m512)(W)))
6722
6726
 
6723
6727
  #define _mm512_maskz_shuffle_ps(U, A, B, M) \
6724
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6725
- (__v16sf)_mm512_shuffle_ps((A), (B), (M)), \
6726
- (__v16sf)_mm512_setzero_ps())
6728
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
6729
+ (__v16sf)_mm512_shuffle_ps((A), (B), (M)), \
6730
+ (__v16sf)_mm512_setzero_ps()))
6727
6731
 
6728
6732
  #define _mm_sqrt_round_sd(A, B, R) \
6729
- (__m128d)__builtin_ia32_sqrtsd_round_mask((__v2df)(__m128d)(A), \
6730
- (__v2df)(__m128d)(B), \
6731
- (__v2df)_mm_setzero_pd(), \
6732
- (__mmask8)-1, (int)(R))
6733
+ ((__m128d)__builtin_ia32_sqrtsd_round_mask((__v2df)(__m128d)(A), \
6734
+ (__v2df)(__m128d)(B), \
6735
+ (__v2df)_mm_setzero_pd(), \
6736
+ (__mmask8)-1, (int)(R)))
6733
6737
 
6734
6738
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
6735
6739
  _mm_mask_sqrt_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
@@ -6742,10 +6746,10 @@ _mm_mask_sqrt_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
6742
6746
  }
6743
6747
 
6744
6748
  #define _mm_mask_sqrt_round_sd(W, U, A, B, R) \
6745
- (__m128d)__builtin_ia32_sqrtsd_round_mask((__v2df)(__m128d)(A), \
6746
- (__v2df)(__m128d)(B), \
6747
- (__v2df)(__m128d)(W), \
6748
- (__mmask8)(U), (int)(R))
6749
+ ((__m128d)__builtin_ia32_sqrtsd_round_mask((__v2df)(__m128d)(A), \
6750
+ (__v2df)(__m128d)(B), \
6751
+ (__v2df)(__m128d)(W), \
6752
+ (__mmask8)(U), (int)(R)))
6749
6753
 
6750
6754
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
6751
6755
  _mm_maskz_sqrt_sd (__mmask8 __U, __m128d __A, __m128d __B)
@@ -6758,16 +6762,16 @@ _mm_maskz_sqrt_sd (__mmask8 __U, __m128d __A, __m128d __B)
6758
6762
  }
6759
6763
 
6760
6764
  #define _mm_maskz_sqrt_round_sd(U, A, B, R) \
6761
- (__m128d)__builtin_ia32_sqrtsd_round_mask((__v2df)(__m128d)(A), \
6762
- (__v2df)(__m128d)(B), \
6763
- (__v2df)_mm_setzero_pd(), \
6764
- (__mmask8)(U), (int)(R))
6765
+ ((__m128d)__builtin_ia32_sqrtsd_round_mask((__v2df)(__m128d)(A), \
6766
+ (__v2df)(__m128d)(B), \
6767
+ (__v2df)_mm_setzero_pd(), \
6768
+ (__mmask8)(U), (int)(R)))
6765
6769
 
6766
6770
  #define _mm_sqrt_round_ss(A, B, R) \
6767
- (__m128)__builtin_ia32_sqrtss_round_mask((__v4sf)(__m128)(A), \
6768
- (__v4sf)(__m128)(B), \
6769
- (__v4sf)_mm_setzero_ps(), \
6770
- (__mmask8)-1, (int)(R))
6771
+ ((__m128)__builtin_ia32_sqrtss_round_mask((__v4sf)(__m128)(A), \
6772
+ (__v4sf)(__m128)(B), \
6773
+ (__v4sf)_mm_setzero_ps(), \
6774
+ (__mmask8)-1, (int)(R)))
6771
6775
 
6772
6776
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
6773
6777
  _mm_mask_sqrt_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
@@ -6780,10 +6784,10 @@ _mm_mask_sqrt_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
6780
6784
  }
6781
6785
 
6782
6786
  #define _mm_mask_sqrt_round_ss(W, U, A, B, R) \
6783
- (__m128)__builtin_ia32_sqrtss_round_mask((__v4sf)(__m128)(A), \
6784
- (__v4sf)(__m128)(B), \
6785
- (__v4sf)(__m128)(W), (__mmask8)(U), \
6786
- (int)(R))
6787
+ ((__m128)__builtin_ia32_sqrtss_round_mask((__v4sf)(__m128)(A), \
6788
+ (__v4sf)(__m128)(B), \
6789
+ (__v4sf)(__m128)(W), (__mmask8)(U), \
6790
+ (int)(R)))
6787
6791
 
6788
6792
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
6789
6793
  _mm_maskz_sqrt_ss (__mmask8 __U, __m128 __A, __m128 __B)
@@ -6796,10 +6800,10 @@ _mm_maskz_sqrt_ss (__mmask8 __U, __m128 __A, __m128 __B)
6796
6800
  }
6797
6801
 
6798
6802
  #define _mm_maskz_sqrt_round_ss(U, A, B, R) \
6799
- (__m128)__builtin_ia32_sqrtss_round_mask((__v4sf)(__m128)(A), \
6800
- (__v4sf)(__m128)(B), \
6801
- (__v4sf)_mm_setzero_ps(), \
6802
- (__mmask8)(U), (int)(R))
6803
+ ((__m128)__builtin_ia32_sqrtss_round_mask((__v4sf)(__m128)(A), \
6804
+ (__v4sf)(__m128)(B), \
6805
+ (__v4sf)_mm_setzero_ps(), \
6806
+ (__mmask8)(U), (int)(R)))
6803
6807
 
6804
6808
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
6805
6809
  _mm512_broadcast_f32x4(__m128 __A)
@@ -7366,183 +7370,183 @@ _mm512_mask_cvtepi64_storeu_epi16 (void *__P, __mmask8 __M, __m512i __A)
7366
7370
  }
7367
7371
 
7368
7372
  #define _mm512_extracti32x4_epi32(A, imm) \
7369
- (__m128i)__builtin_ia32_extracti32x4_mask((__v16si)(__m512i)(A), (int)(imm), \
7370
- (__v4si)_mm_undefined_si128(), \
7371
- (__mmask8)-1)
7373
+ ((__m128i)__builtin_ia32_extracti32x4_mask((__v16si)(__m512i)(A), (int)(imm), \
7374
+ (__v4si)_mm_undefined_si128(), \
7375
+ (__mmask8)-1))
7372
7376
 
7373
7377
  #define _mm512_mask_extracti32x4_epi32(W, U, A, imm) \
7374
- (__m128i)__builtin_ia32_extracti32x4_mask((__v16si)(__m512i)(A), (int)(imm), \
7375
- (__v4si)(__m128i)(W), \
7376
- (__mmask8)(U))
7378
+ ((__m128i)__builtin_ia32_extracti32x4_mask((__v16si)(__m512i)(A), (int)(imm), \
7379
+ (__v4si)(__m128i)(W), \
7380
+ (__mmask8)(U)))
7377
7381
 
7378
7382
  #define _mm512_maskz_extracti32x4_epi32(U, A, imm) \
7379
- (__m128i)__builtin_ia32_extracti32x4_mask((__v16si)(__m512i)(A), (int)(imm), \
7380
- (__v4si)_mm_setzero_si128(), \
7381
- (__mmask8)(U))
7383
+ ((__m128i)__builtin_ia32_extracti32x4_mask((__v16si)(__m512i)(A), (int)(imm), \
7384
+ (__v4si)_mm_setzero_si128(), \
7385
+ (__mmask8)(U)))
7382
7386
 
7383
7387
  #define _mm512_extracti64x4_epi64(A, imm) \
7384
- (__m256i)__builtin_ia32_extracti64x4_mask((__v8di)(__m512i)(A), (int)(imm), \
7385
- (__v4di)_mm256_undefined_si256(), \
7386
- (__mmask8)-1)
7388
+ ((__m256i)__builtin_ia32_extracti64x4_mask((__v8di)(__m512i)(A), (int)(imm), \
7389
+ (__v4di)_mm256_undefined_si256(), \
7390
+ (__mmask8)-1))
7387
7391
 
7388
7392
  #define _mm512_mask_extracti64x4_epi64(W, U, A, imm) \
7389
- (__m256i)__builtin_ia32_extracti64x4_mask((__v8di)(__m512i)(A), (int)(imm), \
7390
- (__v4di)(__m256i)(W), \
7391
- (__mmask8)(U))
7393
+ ((__m256i)__builtin_ia32_extracti64x4_mask((__v8di)(__m512i)(A), (int)(imm), \
7394
+ (__v4di)(__m256i)(W), \
7395
+ (__mmask8)(U)))
7392
7396
 
7393
7397
  #define _mm512_maskz_extracti64x4_epi64(U, A, imm) \
7394
- (__m256i)__builtin_ia32_extracti64x4_mask((__v8di)(__m512i)(A), (int)(imm), \
7395
- (__v4di)_mm256_setzero_si256(), \
7396
- (__mmask8)(U))
7398
+ ((__m256i)__builtin_ia32_extracti64x4_mask((__v8di)(__m512i)(A), (int)(imm), \
7399
+ (__v4di)_mm256_setzero_si256(), \
7400
+ (__mmask8)(U)))
7397
7401
 
7398
7402
  #define _mm512_insertf64x4(A, B, imm) \
7399
- (__m512d)__builtin_ia32_insertf64x4((__v8df)(__m512d)(A), \
7400
- (__v4df)(__m256d)(B), (int)(imm))
7403
+ ((__m512d)__builtin_ia32_insertf64x4((__v8df)(__m512d)(A), \
7404
+ (__v4df)(__m256d)(B), (int)(imm)))
7401
7405
 
7402
7406
  #define _mm512_mask_insertf64x4(W, U, A, B, imm) \
7403
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
7404
- (__v8df)_mm512_insertf64x4((A), (B), (imm)), \
7405
- (__v8df)(__m512d)(W))
7407
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
7408
+ (__v8df)_mm512_insertf64x4((A), (B), (imm)), \
7409
+ (__v8df)(__m512d)(W)))
7406
7410
 
7407
7411
  #define _mm512_maskz_insertf64x4(U, A, B, imm) \
7408
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
7409
- (__v8df)_mm512_insertf64x4((A), (B), (imm)), \
7410
- (__v8df)_mm512_setzero_pd())
7412
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
7413
+ (__v8df)_mm512_insertf64x4((A), (B), (imm)), \
7414
+ (__v8df)_mm512_setzero_pd()))
7411
7415
 
7412
7416
  #define _mm512_inserti64x4(A, B, imm) \
7413
- (__m512i)__builtin_ia32_inserti64x4((__v8di)(__m512i)(A), \
7414
- (__v4di)(__m256i)(B), (int)(imm))
7417
+ ((__m512i)__builtin_ia32_inserti64x4((__v8di)(__m512i)(A), \
7418
+ (__v4di)(__m256i)(B), (int)(imm)))
7415
7419
 
7416
7420
  #define _mm512_mask_inserti64x4(W, U, A, B, imm) \
7417
- (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
7418
- (__v8di)_mm512_inserti64x4((A), (B), (imm)), \
7419
- (__v8di)(__m512i)(W))
7421
+ ((__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
7422
+ (__v8di)_mm512_inserti64x4((A), (B), (imm)), \
7423
+ (__v8di)(__m512i)(W)))
7420
7424
 
7421
7425
  #define _mm512_maskz_inserti64x4(U, A, B, imm) \
7422
- (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
7423
- (__v8di)_mm512_inserti64x4((A), (B), (imm)), \
7424
- (__v8di)_mm512_setzero_si512())
7426
+ ((__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
7427
+ (__v8di)_mm512_inserti64x4((A), (B), (imm)), \
7428
+ (__v8di)_mm512_setzero_si512()))
7425
7429
 
7426
7430
  #define _mm512_insertf32x4(A, B, imm) \
7427
- (__m512)__builtin_ia32_insertf32x4((__v16sf)(__m512)(A), \
7428
- (__v4sf)(__m128)(B), (int)(imm))
7431
+ ((__m512)__builtin_ia32_insertf32x4((__v16sf)(__m512)(A), \
7432
+ (__v4sf)(__m128)(B), (int)(imm)))
7429
7433
 
7430
7434
  #define _mm512_mask_insertf32x4(W, U, A, B, imm) \
7431
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
7432
- (__v16sf)_mm512_insertf32x4((A), (B), (imm)), \
7433
- (__v16sf)(__m512)(W))
7435
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
7436
+ (__v16sf)_mm512_insertf32x4((A), (B), (imm)), \
7437
+ (__v16sf)(__m512)(W)))
7434
7438
 
7435
7439
  #define _mm512_maskz_insertf32x4(U, A, B, imm) \
7436
- (__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
7437
- (__v16sf)_mm512_insertf32x4((A), (B), (imm)), \
7438
- (__v16sf)_mm512_setzero_ps())
7440
+ ((__m512)__builtin_ia32_selectps_512((__mmask16)(U), \
7441
+ (__v16sf)_mm512_insertf32x4((A), (B), (imm)), \
7442
+ (__v16sf)_mm512_setzero_ps()))
7439
7443
 
7440
7444
  #define _mm512_inserti32x4(A, B, imm) \
7441
- (__m512i)__builtin_ia32_inserti32x4((__v16si)(__m512i)(A), \
7442
- (__v4si)(__m128i)(B), (int)(imm))
7445
+ ((__m512i)__builtin_ia32_inserti32x4((__v16si)(__m512i)(A), \
7446
+ (__v4si)(__m128i)(B), (int)(imm)))
7443
7447
 
7444
7448
  #define _mm512_mask_inserti32x4(W, U, A, B, imm) \
7445
- (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
7446
- (__v16si)_mm512_inserti32x4((A), (B), (imm)), \
7447
- (__v16si)(__m512i)(W))
7449
+ ((__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
7450
+ (__v16si)_mm512_inserti32x4((A), (B), (imm)), \
7451
+ (__v16si)(__m512i)(W)))
7448
7452
 
7449
7453
  #define _mm512_maskz_inserti32x4(U, A, B, imm) \
7450
- (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
7451
- (__v16si)_mm512_inserti32x4((A), (B), (imm)), \
7452
- (__v16si)_mm512_setzero_si512())
7454
+ ((__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
7455
+ (__v16si)_mm512_inserti32x4((A), (B), (imm)), \
7456
+ (__v16si)_mm512_setzero_si512()))
7453
7457
 
7454
7458
  #define _mm512_getmant_round_pd(A, B, C, R) \
7455
- (__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
7456
- (int)(((C)<<2) | (B)), \
7457
- (__v8df)_mm512_undefined_pd(), \
7458
- (__mmask8)-1, (int)(R))
7459
+ ((__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
7460
+ (int)(((C)<<2) | (B)), \
7461
+ (__v8df)_mm512_undefined_pd(), \
7462
+ (__mmask8)-1, (int)(R)))
7459
7463
 
7460
7464
  #define _mm512_mask_getmant_round_pd(W, U, A, B, C, R) \
7461
- (__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
7462
- (int)(((C)<<2) | (B)), \
7463
- (__v8df)(__m512d)(W), \
7464
- (__mmask8)(U), (int)(R))
7465
+ ((__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
7466
+ (int)(((C)<<2) | (B)), \
7467
+ (__v8df)(__m512d)(W), \
7468
+ (__mmask8)(U), (int)(R)))
7465
7469
 
7466
7470
  #define _mm512_maskz_getmant_round_pd(U, A, B, C, R) \
7467
- (__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
7468
- (int)(((C)<<2) | (B)), \
7469
- (__v8df)_mm512_setzero_pd(), \
7470
- (__mmask8)(U), (int)(R))
7471
+ ((__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
7472
+ (int)(((C)<<2) | (B)), \
7473
+ (__v8df)_mm512_setzero_pd(), \
7474
+ (__mmask8)(U), (int)(R)))
7471
7475
 
7472
7476
  #define _mm512_getmant_pd(A, B, C) \
7473
- (__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
7474
- (int)(((C)<<2) | (B)), \
7475
- (__v8df)_mm512_setzero_pd(), \
7476
- (__mmask8)-1, \
7477
- _MM_FROUND_CUR_DIRECTION)
7477
+ ((__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
7478
+ (int)(((C)<<2) | (B)), \
7479
+ (__v8df)_mm512_setzero_pd(), \
7480
+ (__mmask8)-1, \
7481
+ _MM_FROUND_CUR_DIRECTION))
7478
7482
 
7479
7483
  #define _mm512_mask_getmant_pd(W, U, A, B, C) \
7480
- (__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
7481
- (int)(((C)<<2) | (B)), \
7482
- (__v8df)(__m512d)(W), \
7483
- (__mmask8)(U), \
7484
- _MM_FROUND_CUR_DIRECTION)
7484
+ ((__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
7485
+ (int)(((C)<<2) | (B)), \
7486
+ (__v8df)(__m512d)(W), \
7487
+ (__mmask8)(U), \
7488
+ _MM_FROUND_CUR_DIRECTION))
7485
7489
 
7486
7490
  #define _mm512_maskz_getmant_pd(U, A, B, C) \
7487
- (__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
7488
- (int)(((C)<<2) | (B)), \
7489
- (__v8df)_mm512_setzero_pd(), \
7490
- (__mmask8)(U), \
7491
- _MM_FROUND_CUR_DIRECTION)
7491
+ ((__m512d)__builtin_ia32_getmantpd512_mask((__v8df)(__m512d)(A), \
7492
+ (int)(((C)<<2) | (B)), \
7493
+ (__v8df)_mm512_setzero_pd(), \
7494
+ (__mmask8)(U), \
7495
+ _MM_FROUND_CUR_DIRECTION))
7492
7496
 
7493
7497
  #define _mm512_getmant_round_ps(A, B, C, R) \
7494
- (__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
7495
- (int)(((C)<<2) | (B)), \
7496
- (__v16sf)_mm512_undefined_ps(), \
7497
- (__mmask16)-1, (int)(R))
7498
+ ((__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
7499
+ (int)(((C)<<2) | (B)), \
7500
+ (__v16sf)_mm512_undefined_ps(), \
7501
+ (__mmask16)-1, (int)(R)))
7498
7502
 
7499
7503
  #define _mm512_mask_getmant_round_ps(W, U, A, B, C, R) \
7500
- (__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
7501
- (int)(((C)<<2) | (B)), \
7502
- (__v16sf)(__m512)(W), \
7503
- (__mmask16)(U), (int)(R))
7504
+ ((__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
7505
+ (int)(((C)<<2) | (B)), \
7506
+ (__v16sf)(__m512)(W), \
7507
+ (__mmask16)(U), (int)(R)))
7504
7508
 
7505
7509
  #define _mm512_maskz_getmant_round_ps(U, A, B, C, R) \
7506
- (__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
7507
- (int)(((C)<<2) | (B)), \
7508
- (__v16sf)_mm512_setzero_ps(), \
7509
- (__mmask16)(U), (int)(R))
7510
+ ((__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
7511
+ (int)(((C)<<2) | (B)), \
7512
+ (__v16sf)_mm512_setzero_ps(), \
7513
+ (__mmask16)(U), (int)(R)))
7510
7514
 
7511
7515
  #define _mm512_getmant_ps(A, B, C) \
7512
- (__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
7513
- (int)(((C)<<2)|(B)), \
7514
- (__v16sf)_mm512_undefined_ps(), \
7515
- (__mmask16)-1, \
7516
- _MM_FROUND_CUR_DIRECTION)
7516
+ ((__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
7517
+ (int)(((C)<<2)|(B)), \
7518
+ (__v16sf)_mm512_undefined_ps(), \
7519
+ (__mmask16)-1, \
7520
+ _MM_FROUND_CUR_DIRECTION))
7517
7521
 
7518
7522
  #define _mm512_mask_getmant_ps(W, U, A, B, C) \
7519
- (__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
7520
- (int)(((C)<<2)|(B)), \
7521
- (__v16sf)(__m512)(W), \
7522
- (__mmask16)(U), \
7523
- _MM_FROUND_CUR_DIRECTION)
7523
+ ((__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
7524
+ (int)(((C)<<2)|(B)), \
7525
+ (__v16sf)(__m512)(W), \
7526
+ (__mmask16)(U), \
7527
+ _MM_FROUND_CUR_DIRECTION))
7524
7528
 
7525
7529
  #define _mm512_maskz_getmant_ps(U, A, B, C) \
7526
- (__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
7527
- (int)(((C)<<2)|(B)), \
7528
- (__v16sf)_mm512_setzero_ps(), \
7529
- (__mmask16)(U), \
7530
- _MM_FROUND_CUR_DIRECTION)
7530
+ ((__m512)__builtin_ia32_getmantps512_mask((__v16sf)(__m512)(A), \
7531
+ (int)(((C)<<2)|(B)), \
7532
+ (__v16sf)_mm512_setzero_ps(), \
7533
+ (__mmask16)(U), \
7534
+ _MM_FROUND_CUR_DIRECTION))
7531
7535
 
7532
7536
  #define _mm512_getexp_round_pd(A, R) \
7533
- (__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
7534
- (__v8df)_mm512_undefined_pd(), \
7535
- (__mmask8)-1, (int)(R))
7537
+ ((__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
7538
+ (__v8df)_mm512_undefined_pd(), \
7539
+ (__mmask8)-1, (int)(R)))
7536
7540
 
7537
7541
  #define _mm512_mask_getexp_round_pd(W, U, A, R) \
7538
- (__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
7539
- (__v8df)(__m512d)(W), \
7540
- (__mmask8)(U), (int)(R))
7542
+ ((__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
7543
+ (__v8df)(__m512d)(W), \
7544
+ (__mmask8)(U), (int)(R)))
7541
7545
 
7542
7546
  #define _mm512_maskz_getexp_round_pd(U, A, R) \
7543
- (__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
7544
- (__v8df)_mm512_setzero_pd(), \
7545
- (__mmask8)(U), (int)(R))
7547
+ ((__m512d)__builtin_ia32_getexppd512_mask((__v8df)(__m512d)(A), \
7548
+ (__v8df)_mm512_setzero_pd(), \
7549
+ (__mmask8)(U), (int)(R)))
7546
7550
 
7547
7551
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
7548
7552
  _mm512_getexp_pd (__m512d __A)
@@ -7572,19 +7576,19 @@ _mm512_maskz_getexp_pd (__mmask8 __U, __m512d __A)
7572
7576
  }
7573
7577
 
7574
7578
  #define _mm512_getexp_round_ps(A, R) \
7575
- (__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
7576
- (__v16sf)_mm512_undefined_ps(), \
7577
- (__mmask16)-1, (int)(R))
7579
+ ((__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
7580
+ (__v16sf)_mm512_undefined_ps(), \
7581
+ (__mmask16)-1, (int)(R)))
7578
7582
 
7579
7583
  #define _mm512_mask_getexp_round_ps(W, U, A, R) \
7580
- (__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
7581
- (__v16sf)(__m512)(W), \
7582
- (__mmask16)(U), (int)(R))
7584
+ ((__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
7585
+ (__v16sf)(__m512)(W), \
7586
+ (__mmask16)(U), (int)(R)))
7583
7587
 
7584
7588
  #define _mm512_maskz_getexp_round_ps(U, A, R) \
7585
- (__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
7586
- (__v16sf)_mm512_setzero_ps(), \
7587
- (__mmask16)(U), (int)(R))
7589
+ ((__m512)__builtin_ia32_getexpps512_mask((__v16sf)(__m512)(A), \
7590
+ (__v16sf)_mm512_setzero_ps(), \
7591
+ (__mmask16)(U), (int)(R)))
7588
7592
 
7589
7593
  static __inline__ __m512 __DEFAULT_FN_ATTRS512
7590
7594
  _mm512_getexp_ps (__m512 __A)
@@ -7614,100 +7618,100 @@ _mm512_maskz_getexp_ps (__mmask16 __U, __m512 __A)
7614
7618
  }
7615
7619
 
7616
7620
  #define _mm512_i64gather_ps(index, addr, scale) \
7617
- (__m256)__builtin_ia32_gatherdiv16sf((__v8sf)_mm256_undefined_ps(), \
7618
- (void const *)(addr), \
7619
- (__v8di)(__m512i)(index), (__mmask8)-1, \
7620
- (int)(scale))
7621
+ ((__m256)__builtin_ia32_gatherdiv16sf((__v8sf)_mm256_undefined_ps(), \
7622
+ (void const *)(addr), \
7623
+ (__v8di)(__m512i)(index), (__mmask8)-1, \
7624
+ (int)(scale)))
7621
7625
 
7622
7626
  #define _mm512_mask_i64gather_ps(v1_old, mask, index, addr, scale) \
7623
- (__m256)__builtin_ia32_gatherdiv16sf((__v8sf)(__m256)(v1_old),\
7624
- (void const *)(addr), \
7625
- (__v8di)(__m512i)(index), \
7626
- (__mmask8)(mask), (int)(scale))
7627
-
7628
- #define _mm512_i64gather_epi32(index, addr, scale) \
7629
- (__m256i)__builtin_ia32_gatherdiv16si((__v8si)_mm256_undefined_si256(), \
7627
+ ((__m256)__builtin_ia32_gatherdiv16sf((__v8sf)(__m256)(v1_old),\
7630
7628
  (void const *)(addr), \
7631
7629
  (__v8di)(__m512i)(index), \
7632
- (__mmask8)-1, (int)(scale))
7630
+ (__mmask8)(mask), (int)(scale)))
7631
+
7632
+ #define _mm512_i64gather_epi32(index, addr, scale) \
7633
+ ((__m256i)__builtin_ia32_gatherdiv16si((__v8si)_mm256_undefined_si256(), \
7634
+ (void const *)(addr), \
7635
+ (__v8di)(__m512i)(index), \
7636
+ (__mmask8)-1, (int)(scale)))
7633
7637
 
7634
7638
  #define _mm512_mask_i64gather_epi32(v1_old, mask, index, addr, scale) \
7635
- (__m256i)__builtin_ia32_gatherdiv16si((__v8si)(__m256i)(v1_old), \
7636
- (void const *)(addr), \
7637
- (__v8di)(__m512i)(index), \
7638
- (__mmask8)(mask), (int)(scale))
7639
+ ((__m256i)__builtin_ia32_gatherdiv16si((__v8si)(__m256i)(v1_old), \
7640
+ (void const *)(addr), \
7641
+ (__v8di)(__m512i)(index), \
7642
+ (__mmask8)(mask), (int)(scale)))
7639
7643
 
7640
7644
  #define _mm512_i64gather_pd(index, addr, scale) \
7641
- (__m512d)__builtin_ia32_gatherdiv8df((__v8df)_mm512_undefined_pd(), \
7642
- (void const *)(addr), \
7643
- (__v8di)(__m512i)(index), (__mmask8)-1, \
7644
- (int)(scale))
7645
+ ((__m512d)__builtin_ia32_gatherdiv8df((__v8df)_mm512_undefined_pd(), \
7646
+ (void const *)(addr), \
7647
+ (__v8di)(__m512i)(index), (__mmask8)-1, \
7648
+ (int)(scale)))
7645
7649
 
7646
7650
  #define _mm512_mask_i64gather_pd(v1_old, mask, index, addr, scale) \
7647
- (__m512d)__builtin_ia32_gatherdiv8df((__v8df)(__m512d)(v1_old), \
7648
- (void const *)(addr), \
7649
- (__v8di)(__m512i)(index), \
7650
- (__mmask8)(mask), (int)(scale))
7651
+ ((__m512d)__builtin_ia32_gatherdiv8df((__v8df)(__m512d)(v1_old), \
7652
+ (void const *)(addr), \
7653
+ (__v8di)(__m512i)(index), \
7654
+ (__mmask8)(mask), (int)(scale)))
7651
7655
 
7652
7656
  #define _mm512_i64gather_epi64(index, addr, scale) \
7653
- (__m512i)__builtin_ia32_gatherdiv8di((__v8di)_mm512_undefined_epi32(), \
7654
- (void const *)(addr), \
7655
- (__v8di)(__m512i)(index), (__mmask8)-1, \
7656
- (int)(scale))
7657
+ ((__m512i)__builtin_ia32_gatherdiv8di((__v8di)_mm512_undefined_epi32(), \
7658
+ (void const *)(addr), \
7659
+ (__v8di)(__m512i)(index), (__mmask8)-1, \
7660
+ (int)(scale)))
7657
7661
 
7658
7662
  #define _mm512_mask_i64gather_epi64(v1_old, mask, index, addr, scale) \
7659
- (__m512i)__builtin_ia32_gatherdiv8di((__v8di)(__m512i)(v1_old), \
7660
- (void const *)(addr), \
7661
- (__v8di)(__m512i)(index), \
7662
- (__mmask8)(mask), (int)(scale))
7663
+ ((__m512i)__builtin_ia32_gatherdiv8di((__v8di)(__m512i)(v1_old), \
7664
+ (void const *)(addr), \
7665
+ (__v8di)(__m512i)(index), \
7666
+ (__mmask8)(mask), (int)(scale)))
7663
7667
 
7664
7668
  #define _mm512_i32gather_ps(index, addr, scale) \
7665
- (__m512)__builtin_ia32_gathersiv16sf((__v16sf)_mm512_undefined_ps(), \
7666
- (void const *)(addr), \
7667
- (__v16si)(__m512)(index), \
7668
- (__mmask16)-1, (int)(scale))
7669
+ ((__m512)__builtin_ia32_gathersiv16sf((__v16sf)_mm512_undefined_ps(), \
7670
+ (void const *)(addr), \
7671
+ (__v16si)(__m512)(index), \
7672
+ (__mmask16)-1, (int)(scale)))
7669
7673
 
7670
7674
  #define _mm512_mask_i32gather_ps(v1_old, mask, index, addr, scale) \
7671
- (__m512)__builtin_ia32_gathersiv16sf((__v16sf)(__m512)(v1_old), \
7672
- (void const *)(addr), \
7673
- (__v16si)(__m512)(index), \
7674
- (__mmask16)(mask), (int)(scale))
7675
+ ((__m512)__builtin_ia32_gathersiv16sf((__v16sf)(__m512)(v1_old), \
7676
+ (void const *)(addr), \
7677
+ (__v16si)(__m512)(index), \
7678
+ (__mmask16)(mask), (int)(scale)))
7675
7679
 
7676
7680
  #define _mm512_i32gather_epi32(index, addr, scale) \
7677
- (__m512i)__builtin_ia32_gathersiv16si((__v16si)_mm512_undefined_epi32(), \
7678
- (void const *)(addr), \
7679
- (__v16si)(__m512i)(index), \
7680
- (__mmask16)-1, (int)(scale))
7681
+ ((__m512i)__builtin_ia32_gathersiv16si((__v16si)_mm512_undefined_epi32(), \
7682
+ (void const *)(addr), \
7683
+ (__v16si)(__m512i)(index), \
7684
+ (__mmask16)-1, (int)(scale)))
7681
7685
 
7682
7686
  #define _mm512_mask_i32gather_epi32(v1_old, mask, index, addr, scale) \
7683
- (__m512i)__builtin_ia32_gathersiv16si((__v16si)(__m512i)(v1_old), \
7684
- (void const *)(addr), \
7685
- (__v16si)(__m512i)(index), \
7686
- (__mmask16)(mask), (int)(scale))
7687
+ ((__m512i)__builtin_ia32_gathersiv16si((__v16si)(__m512i)(v1_old), \
7688
+ (void const *)(addr), \
7689
+ (__v16si)(__m512i)(index), \
7690
+ (__mmask16)(mask), (int)(scale)))
7687
7691
 
7688
7692
  #define _mm512_i32gather_pd(index, addr, scale) \
7689
- (__m512d)__builtin_ia32_gathersiv8df((__v8df)_mm512_undefined_pd(), \
7690
- (void const *)(addr), \
7691
- (__v8si)(__m256i)(index), (__mmask8)-1, \
7692
- (int)(scale))
7693
+ ((__m512d)__builtin_ia32_gathersiv8df((__v8df)_mm512_undefined_pd(), \
7694
+ (void const *)(addr), \
7695
+ (__v8si)(__m256i)(index), (__mmask8)-1, \
7696
+ (int)(scale)))
7693
7697
 
7694
7698
  #define _mm512_mask_i32gather_pd(v1_old, mask, index, addr, scale) \
7695
- (__m512d)__builtin_ia32_gathersiv8df((__v8df)(__m512d)(v1_old), \
7696
- (void const *)(addr), \
7697
- (__v8si)(__m256i)(index), \
7698
- (__mmask8)(mask), (int)(scale))
7699
+ ((__m512d)__builtin_ia32_gathersiv8df((__v8df)(__m512d)(v1_old), \
7700
+ (void const *)(addr), \
7701
+ (__v8si)(__m256i)(index), \
7702
+ (__mmask8)(mask), (int)(scale)))
7699
7703
 
7700
7704
  #define _mm512_i32gather_epi64(index, addr, scale) \
7701
- (__m512i)__builtin_ia32_gathersiv8di((__v8di)_mm512_undefined_epi32(), \
7702
- (void const *)(addr), \
7703
- (__v8si)(__m256i)(index), (__mmask8)-1, \
7704
- (int)(scale))
7705
+ ((__m512i)__builtin_ia32_gathersiv8di((__v8di)_mm512_undefined_epi32(), \
7706
+ (void const *)(addr), \
7707
+ (__v8si)(__m256i)(index), (__mmask8)-1, \
7708
+ (int)(scale)))
7705
7709
 
7706
7710
  #define _mm512_mask_i32gather_epi64(v1_old, mask, index, addr, scale) \
7707
- (__m512i)__builtin_ia32_gathersiv8di((__v8di)(__m512i)(v1_old), \
7708
- (void const *)(addr), \
7709
- (__v8si)(__m256i)(index), \
7710
- (__mmask8)(mask), (int)(scale))
7711
+ ((__m512i)__builtin_ia32_gathersiv8di((__v8di)(__m512i)(v1_old), \
7712
+ (void const *)(addr), \
7713
+ (__v8si)(__m256i)(index), \
7714
+ (__mmask8)(mask), (int)(scale)))
7711
7715
 
7712
7716
  #define _mm512_i64scatter_ps(addr, index, v1, scale) \
7713
7717
  __builtin_ia32_scatterdiv16sf((void *)(addr), (__mmask8)-1, \
@@ -7800,16 +7804,16 @@ _mm_mask_fmadd_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
7800
7804
  }
7801
7805
 
7802
7806
  #define _mm_fmadd_round_ss(A, B, C, R) \
7803
- (__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(A), \
7804
- (__v4sf)(__m128)(B), \
7805
- (__v4sf)(__m128)(C), (__mmask8)-1, \
7806
- (int)(R))
7807
+ ((__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(A), \
7808
+ (__v4sf)(__m128)(B), \
7809
+ (__v4sf)(__m128)(C), (__mmask8)-1, \
7810
+ (int)(R)))
7807
7811
 
7808
7812
  #define _mm_mask_fmadd_round_ss(W, U, A, B, R) \
7809
- (__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(W), \
7810
- (__v4sf)(__m128)(A), \
7811
- (__v4sf)(__m128)(B), (__mmask8)(U), \
7812
- (int)(R))
7813
+ ((__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(W), \
7814
+ (__v4sf)(__m128)(A), \
7815
+ (__v4sf)(__m128)(B), (__mmask8)(U), \
7816
+ (int)(R)))
7813
7817
 
7814
7818
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
7815
7819
  _mm_maskz_fmadd_ss (__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
@@ -7822,10 +7826,10 @@ _mm_maskz_fmadd_ss (__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
7822
7826
  }
7823
7827
 
7824
7828
  #define _mm_maskz_fmadd_round_ss(U, A, B, C, R) \
7825
- (__m128)__builtin_ia32_vfmaddss3_maskz((__v4sf)(__m128)(A), \
7826
- (__v4sf)(__m128)(B), \
7827
- (__v4sf)(__m128)(C), (__mmask8)(U), \
7828
- (int)(R))
7829
+ ((__m128)__builtin_ia32_vfmaddss3_maskz((__v4sf)(__m128)(A), \
7830
+ (__v4sf)(__m128)(B), \
7831
+ (__v4sf)(__m128)(C), (__mmask8)(U), \
7832
+ (int)(R)))
7829
7833
 
7830
7834
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
7831
7835
  _mm_mask3_fmadd_ss (__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
@@ -7838,10 +7842,10 @@ _mm_mask3_fmadd_ss (__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
7838
7842
  }
7839
7843
 
7840
7844
  #define _mm_mask3_fmadd_round_ss(W, X, Y, U, R) \
7841
- (__m128)__builtin_ia32_vfmaddss3_mask3((__v4sf)(__m128)(W), \
7842
- (__v4sf)(__m128)(X), \
7843
- (__v4sf)(__m128)(Y), (__mmask8)(U), \
7844
- (int)(R))
7845
+ ((__m128)__builtin_ia32_vfmaddss3_mask3((__v4sf)(__m128)(W), \
7846
+ (__v4sf)(__m128)(X), \
7847
+ (__v4sf)(__m128)(Y), (__mmask8)(U), \
7848
+ (int)(R)))
7845
7849
 
7846
7850
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
7847
7851
  _mm_mask_fmsub_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
@@ -7854,16 +7858,16 @@ _mm_mask_fmsub_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
7854
7858
  }
7855
7859
 
7856
7860
  #define _mm_fmsub_round_ss(A, B, C, R) \
7857
- (__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(A), \
7858
- (__v4sf)(__m128)(B), \
7859
- -(__v4sf)(__m128)(C), (__mmask8)-1, \
7860
- (int)(R))
7861
+ ((__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(A), \
7862
+ (__v4sf)(__m128)(B), \
7863
+ -(__v4sf)(__m128)(C), (__mmask8)-1, \
7864
+ (int)(R)))
7861
7865
 
7862
7866
  #define _mm_mask_fmsub_round_ss(W, U, A, B, R) \
7863
- (__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(W), \
7864
- (__v4sf)(__m128)(A), \
7865
- -(__v4sf)(__m128)(B), (__mmask8)(U), \
7866
- (int)(R))
7867
+ ((__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(W), \
7868
+ (__v4sf)(__m128)(A), \
7869
+ -(__v4sf)(__m128)(B), (__mmask8)(U), \
7870
+ (int)(R)))
7867
7871
 
7868
7872
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
7869
7873
  _mm_maskz_fmsub_ss (__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
@@ -7876,10 +7880,10 @@ _mm_maskz_fmsub_ss (__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
7876
7880
  }
7877
7881
 
7878
7882
  #define _mm_maskz_fmsub_round_ss(U, A, B, C, R) \
7879
- (__m128)__builtin_ia32_vfmaddss3_maskz((__v4sf)(__m128)(A), \
7880
- (__v4sf)(__m128)(B), \
7881
- -(__v4sf)(__m128)(C), (__mmask8)(U), \
7882
- (int)(R))
7883
+ ((__m128)__builtin_ia32_vfmaddss3_maskz((__v4sf)(__m128)(A), \
7884
+ (__v4sf)(__m128)(B), \
7885
+ -(__v4sf)(__m128)(C), (__mmask8)(U), \
7886
+ (int)(R)))
7883
7887
 
7884
7888
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
7885
7889
  _mm_mask3_fmsub_ss (__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
@@ -7892,10 +7896,10 @@ _mm_mask3_fmsub_ss (__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
7892
7896
  }
7893
7897
 
7894
7898
  #define _mm_mask3_fmsub_round_ss(W, X, Y, U, R) \
7895
- (__m128)__builtin_ia32_vfmsubss3_mask3((__v4sf)(__m128)(W), \
7896
- (__v4sf)(__m128)(X), \
7897
- (__v4sf)(__m128)(Y), (__mmask8)(U), \
7898
- (int)(R))
7899
+ ((__m128)__builtin_ia32_vfmsubss3_mask3((__v4sf)(__m128)(W), \
7900
+ (__v4sf)(__m128)(X), \
7901
+ (__v4sf)(__m128)(Y), (__mmask8)(U), \
7902
+ (int)(R)))
7899
7903
 
7900
7904
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
7901
7905
  _mm_mask_fnmadd_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
@@ -7908,16 +7912,16 @@ _mm_mask_fnmadd_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
7908
7912
  }
7909
7913
 
7910
7914
  #define _mm_fnmadd_round_ss(A, B, C, R) \
7911
- (__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(A), \
7912
- -(__v4sf)(__m128)(B), \
7913
- (__v4sf)(__m128)(C), (__mmask8)-1, \
7914
- (int)(R))
7915
+ ((__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(A), \
7916
+ -(__v4sf)(__m128)(B), \
7917
+ (__v4sf)(__m128)(C), (__mmask8)-1, \
7918
+ (int)(R)))
7915
7919
 
7916
7920
  #define _mm_mask_fnmadd_round_ss(W, U, A, B, R) \
7917
- (__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(W), \
7918
- -(__v4sf)(__m128)(A), \
7919
- (__v4sf)(__m128)(B), (__mmask8)(U), \
7920
- (int)(R))
7921
+ ((__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(W), \
7922
+ -(__v4sf)(__m128)(A), \
7923
+ (__v4sf)(__m128)(B), (__mmask8)(U), \
7924
+ (int)(R)))
7921
7925
 
7922
7926
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
7923
7927
  _mm_maskz_fnmadd_ss (__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
@@ -7930,10 +7934,10 @@ _mm_maskz_fnmadd_ss (__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
7930
7934
  }
7931
7935
 
7932
7936
  #define _mm_maskz_fnmadd_round_ss(U, A, B, C, R) \
7933
- (__m128)__builtin_ia32_vfmaddss3_maskz((__v4sf)(__m128)(A), \
7934
- -(__v4sf)(__m128)(B), \
7935
- (__v4sf)(__m128)(C), (__mmask8)(U), \
7936
- (int)(R))
7937
+ ((__m128)__builtin_ia32_vfmaddss3_maskz((__v4sf)(__m128)(A), \
7938
+ -(__v4sf)(__m128)(B), \
7939
+ (__v4sf)(__m128)(C), (__mmask8)(U), \
7940
+ (int)(R)))
7937
7941
 
7938
7942
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
7939
7943
  _mm_mask3_fnmadd_ss (__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
@@ -7946,10 +7950,10 @@ _mm_mask3_fnmadd_ss (__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
7946
7950
  }
7947
7951
 
7948
7952
  #define _mm_mask3_fnmadd_round_ss(W, X, Y, U, R) \
7949
- (__m128)__builtin_ia32_vfmaddss3_mask3((__v4sf)(__m128)(W), \
7950
- -(__v4sf)(__m128)(X), \
7951
- (__v4sf)(__m128)(Y), (__mmask8)(U), \
7952
- (int)(R))
7953
+ ((__m128)__builtin_ia32_vfmaddss3_mask3((__v4sf)(__m128)(W), \
7954
+ -(__v4sf)(__m128)(X), \
7955
+ (__v4sf)(__m128)(Y), (__mmask8)(U), \
7956
+ (int)(R)))
7953
7957
 
7954
7958
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
7955
7959
  _mm_mask_fnmsub_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
@@ -7962,16 +7966,16 @@ _mm_mask_fnmsub_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128 __B)
7962
7966
  }
7963
7967
 
7964
7968
  #define _mm_fnmsub_round_ss(A, B, C, R) \
7965
- (__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(A), \
7966
- -(__v4sf)(__m128)(B), \
7967
- -(__v4sf)(__m128)(C), (__mmask8)-1, \
7968
- (int)(R))
7969
+ ((__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(A), \
7970
+ -(__v4sf)(__m128)(B), \
7971
+ -(__v4sf)(__m128)(C), (__mmask8)-1, \
7972
+ (int)(R)))
7969
7973
 
7970
7974
  #define _mm_mask_fnmsub_round_ss(W, U, A, B, R) \
7971
- (__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(W), \
7972
- -(__v4sf)(__m128)(A), \
7973
- -(__v4sf)(__m128)(B), (__mmask8)(U), \
7974
- (int)(R))
7975
+ ((__m128)__builtin_ia32_vfmaddss3_mask((__v4sf)(__m128)(W), \
7976
+ -(__v4sf)(__m128)(A), \
7977
+ -(__v4sf)(__m128)(B), (__mmask8)(U), \
7978
+ (int)(R)))
7975
7979
 
7976
7980
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
7977
7981
  _mm_maskz_fnmsub_ss (__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
@@ -7984,10 +7988,10 @@ _mm_maskz_fnmsub_ss (__mmask8 __U, __m128 __A, __m128 __B, __m128 __C)
7984
7988
  }
7985
7989
 
7986
7990
  #define _mm_maskz_fnmsub_round_ss(U, A, B, C, R) \
7987
- (__m128)__builtin_ia32_vfmaddss3_maskz((__v4sf)(__m128)(A), \
7988
- -(__v4sf)(__m128)(B), \
7989
- -(__v4sf)(__m128)(C), (__mmask8)(U), \
7990
- (int)(R))
7991
+ ((__m128)__builtin_ia32_vfmaddss3_maskz((__v4sf)(__m128)(A), \
7992
+ -(__v4sf)(__m128)(B), \
7993
+ -(__v4sf)(__m128)(C), (__mmask8)(U), \
7994
+ (int)(R)))
7991
7995
 
7992
7996
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
7993
7997
  _mm_mask3_fnmsub_ss (__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
@@ -8000,10 +8004,10 @@ _mm_mask3_fnmsub_ss (__m128 __W, __m128 __X, __m128 __Y, __mmask8 __U)
8000
8004
  }
8001
8005
 
8002
8006
  #define _mm_mask3_fnmsub_round_ss(W, X, Y, U, R) \
8003
- (__m128)__builtin_ia32_vfmsubss3_mask3((__v4sf)(__m128)(W), \
8004
- -(__v4sf)(__m128)(X), \
8005
- (__v4sf)(__m128)(Y), (__mmask8)(U), \
8006
- (int)(R))
8007
+ ((__m128)__builtin_ia32_vfmsubss3_mask3((__v4sf)(__m128)(W), \
8008
+ -(__v4sf)(__m128)(X), \
8009
+ (__v4sf)(__m128)(Y), (__mmask8)(U), \
8010
+ (int)(R)))
8007
8011
 
8008
8012
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
8009
8013
  _mm_mask_fmadd_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
@@ -8016,16 +8020,16 @@ _mm_mask_fmadd_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
8016
8020
  }
8017
8021
 
8018
8022
  #define _mm_fmadd_round_sd(A, B, C, R) \
8019
- (__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(A), \
8020
- (__v2df)(__m128d)(B), \
8021
- (__v2df)(__m128d)(C), (__mmask8)-1, \
8022
- (int)(R))
8023
+ ((__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(A), \
8024
+ (__v2df)(__m128d)(B), \
8025
+ (__v2df)(__m128d)(C), (__mmask8)-1, \
8026
+ (int)(R)))
8023
8027
 
8024
8028
  #define _mm_mask_fmadd_round_sd(W, U, A, B, R) \
8025
- (__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(W), \
8026
- (__v2df)(__m128d)(A), \
8027
- (__v2df)(__m128d)(B), (__mmask8)(U), \
8028
- (int)(R))
8029
+ ((__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(W), \
8030
+ (__v2df)(__m128d)(A), \
8031
+ (__v2df)(__m128d)(B), (__mmask8)(U), \
8032
+ (int)(R)))
8029
8033
 
8030
8034
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
8031
8035
  _mm_maskz_fmadd_sd (__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
@@ -8038,10 +8042,10 @@ _mm_maskz_fmadd_sd (__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
8038
8042
  }
8039
8043
 
8040
8044
  #define _mm_maskz_fmadd_round_sd(U, A, B, C, R) \
8041
- (__m128d)__builtin_ia32_vfmaddsd3_maskz((__v2df)(__m128d)(A), \
8042
- (__v2df)(__m128d)(B), \
8043
- (__v2df)(__m128d)(C), (__mmask8)(U), \
8044
- (int)(R))
8045
+ ((__m128d)__builtin_ia32_vfmaddsd3_maskz((__v2df)(__m128d)(A), \
8046
+ (__v2df)(__m128d)(B), \
8047
+ (__v2df)(__m128d)(C), (__mmask8)(U), \
8048
+ (int)(R)))
8045
8049
 
8046
8050
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
8047
8051
  _mm_mask3_fmadd_sd (__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
@@ -8054,10 +8058,10 @@ _mm_mask3_fmadd_sd (__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
8054
8058
  }
8055
8059
 
8056
8060
  #define _mm_mask3_fmadd_round_sd(W, X, Y, U, R) \
8057
- (__m128d)__builtin_ia32_vfmaddsd3_mask3((__v2df)(__m128d)(W), \
8058
- (__v2df)(__m128d)(X), \
8059
- (__v2df)(__m128d)(Y), (__mmask8)(U), \
8060
- (int)(R))
8061
+ ((__m128d)__builtin_ia32_vfmaddsd3_mask3((__v2df)(__m128d)(W), \
8062
+ (__v2df)(__m128d)(X), \
8063
+ (__v2df)(__m128d)(Y), (__mmask8)(U), \
8064
+ (int)(R)))
8061
8065
 
8062
8066
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
8063
8067
  _mm_mask_fmsub_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
@@ -8070,16 +8074,16 @@ _mm_mask_fmsub_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
8070
8074
  }
8071
8075
 
8072
8076
  #define _mm_fmsub_round_sd(A, B, C, R) \
8073
- (__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(A), \
8074
- (__v2df)(__m128d)(B), \
8075
- -(__v2df)(__m128d)(C), (__mmask8)-1, \
8076
- (int)(R))
8077
+ ((__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(A), \
8078
+ (__v2df)(__m128d)(B), \
8079
+ -(__v2df)(__m128d)(C), (__mmask8)-1, \
8080
+ (int)(R)))
8077
8081
 
8078
8082
  #define _mm_mask_fmsub_round_sd(W, U, A, B, R) \
8079
- (__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(W), \
8080
- (__v2df)(__m128d)(A), \
8081
- -(__v2df)(__m128d)(B), (__mmask8)(U), \
8082
- (int)(R))
8083
+ ((__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(W), \
8084
+ (__v2df)(__m128d)(A), \
8085
+ -(__v2df)(__m128d)(B), (__mmask8)(U), \
8086
+ (int)(R)))
8083
8087
 
8084
8088
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
8085
8089
  _mm_maskz_fmsub_sd (__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
@@ -8092,10 +8096,10 @@ _mm_maskz_fmsub_sd (__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
8092
8096
  }
8093
8097
 
8094
8098
  #define _mm_maskz_fmsub_round_sd(U, A, B, C, R) \
8095
- (__m128d)__builtin_ia32_vfmaddsd3_maskz((__v2df)(__m128d)(A), \
8096
- (__v2df)(__m128d)(B), \
8097
- -(__v2df)(__m128d)(C), \
8098
- (__mmask8)(U), (int)(R))
8099
+ ((__m128d)__builtin_ia32_vfmaddsd3_maskz((__v2df)(__m128d)(A), \
8100
+ (__v2df)(__m128d)(B), \
8101
+ -(__v2df)(__m128d)(C), \
8102
+ (__mmask8)(U), (int)(R)))
8099
8103
 
8100
8104
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
8101
8105
  _mm_mask3_fmsub_sd (__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
@@ -8108,10 +8112,10 @@ _mm_mask3_fmsub_sd (__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
8108
8112
  }
8109
8113
 
8110
8114
  #define _mm_mask3_fmsub_round_sd(W, X, Y, U, R) \
8111
- (__m128d)__builtin_ia32_vfmsubsd3_mask3((__v2df)(__m128d)(W), \
8112
- (__v2df)(__m128d)(X), \
8113
- (__v2df)(__m128d)(Y), \
8114
- (__mmask8)(U), (int)(R))
8115
+ ((__m128d)__builtin_ia32_vfmsubsd3_mask3((__v2df)(__m128d)(W), \
8116
+ (__v2df)(__m128d)(X), \
8117
+ (__v2df)(__m128d)(Y), \
8118
+ (__mmask8)(U), (int)(R)))
8115
8119
 
8116
8120
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
8117
8121
  _mm_mask_fnmadd_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
@@ -8124,16 +8128,16 @@ _mm_mask_fnmadd_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
8124
8128
  }
8125
8129
 
8126
8130
  #define _mm_fnmadd_round_sd(A, B, C, R) \
8127
- (__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(A), \
8128
- -(__v2df)(__m128d)(B), \
8129
- (__v2df)(__m128d)(C), (__mmask8)-1, \
8130
- (int)(R))
8131
+ ((__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(A), \
8132
+ -(__v2df)(__m128d)(B), \
8133
+ (__v2df)(__m128d)(C), (__mmask8)-1, \
8134
+ (int)(R)))
8131
8135
 
8132
8136
  #define _mm_mask_fnmadd_round_sd(W, U, A, B, R) \
8133
- (__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(W), \
8134
- -(__v2df)(__m128d)(A), \
8135
- (__v2df)(__m128d)(B), (__mmask8)(U), \
8136
- (int)(R))
8137
+ ((__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(W), \
8138
+ -(__v2df)(__m128d)(A), \
8139
+ (__v2df)(__m128d)(B), (__mmask8)(U), \
8140
+ (int)(R)))
8137
8141
 
8138
8142
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
8139
8143
  _mm_maskz_fnmadd_sd (__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
@@ -8146,10 +8150,10 @@ _mm_maskz_fnmadd_sd (__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
8146
8150
  }
8147
8151
 
8148
8152
  #define _mm_maskz_fnmadd_round_sd(U, A, B, C, R) \
8149
- (__m128d)__builtin_ia32_vfmaddsd3_maskz((__v2df)(__m128d)(A), \
8150
- -(__v2df)(__m128d)(B), \
8151
- (__v2df)(__m128d)(C), (__mmask8)(U), \
8152
- (int)(R))
8153
+ ((__m128d)__builtin_ia32_vfmaddsd3_maskz((__v2df)(__m128d)(A), \
8154
+ -(__v2df)(__m128d)(B), \
8155
+ (__v2df)(__m128d)(C), (__mmask8)(U), \
8156
+ (int)(R)))
8153
8157
 
8154
8158
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
8155
8159
  _mm_mask3_fnmadd_sd (__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
@@ -8162,10 +8166,10 @@ _mm_mask3_fnmadd_sd (__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
8162
8166
  }
8163
8167
 
8164
8168
  #define _mm_mask3_fnmadd_round_sd(W, X, Y, U, R) \
8165
- (__m128d)__builtin_ia32_vfmaddsd3_mask3((__v2df)(__m128d)(W), \
8166
- -(__v2df)(__m128d)(X), \
8167
- (__v2df)(__m128d)(Y), (__mmask8)(U), \
8168
- (int)(R))
8169
+ ((__m128d)__builtin_ia32_vfmaddsd3_mask3((__v2df)(__m128d)(W), \
8170
+ -(__v2df)(__m128d)(X), \
8171
+ (__v2df)(__m128d)(Y), (__mmask8)(U), \
8172
+ (int)(R)))
8169
8173
 
8170
8174
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
8171
8175
  _mm_mask_fnmsub_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
@@ -8178,16 +8182,16 @@ _mm_mask_fnmsub_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128d __B)
8178
8182
  }
8179
8183
 
8180
8184
  #define _mm_fnmsub_round_sd(A, B, C, R) \
8181
- (__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(A), \
8182
- -(__v2df)(__m128d)(B), \
8183
- -(__v2df)(__m128d)(C), (__mmask8)-1, \
8184
- (int)(R))
8185
+ ((__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(A), \
8186
+ -(__v2df)(__m128d)(B), \
8187
+ -(__v2df)(__m128d)(C), (__mmask8)-1, \
8188
+ (int)(R)))
8185
8189
 
8186
8190
  #define _mm_mask_fnmsub_round_sd(W, U, A, B, R) \
8187
- (__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(W), \
8188
- -(__v2df)(__m128d)(A), \
8189
- -(__v2df)(__m128d)(B), (__mmask8)(U), \
8190
- (int)(R))
8191
+ ((__m128d)__builtin_ia32_vfmaddsd3_mask((__v2df)(__m128d)(W), \
8192
+ -(__v2df)(__m128d)(A), \
8193
+ -(__v2df)(__m128d)(B), (__mmask8)(U), \
8194
+ (int)(R)))
8191
8195
 
8192
8196
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
8193
8197
  _mm_maskz_fnmsub_sd (__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
@@ -8200,11 +8204,11 @@ _mm_maskz_fnmsub_sd (__mmask8 __U, __m128d __A, __m128d __B, __m128d __C)
8200
8204
  }
8201
8205
 
8202
8206
  #define _mm_maskz_fnmsub_round_sd(U, A, B, C, R) \
8203
- (__m128d)__builtin_ia32_vfmaddsd3_maskz((__v2df)(__m128d)(A), \
8204
- -(__v2df)(__m128d)(B), \
8205
- -(__v2df)(__m128d)(C), \
8206
- (__mmask8)(U), \
8207
- (int)(R))
8207
+ ((__m128d)__builtin_ia32_vfmaddsd3_maskz((__v2df)(__m128d)(A), \
8208
+ -(__v2df)(__m128d)(B), \
8209
+ -(__v2df)(__m128d)(C), \
8210
+ (__mmask8)(U), \
8211
+ (int)(R)))
8208
8212
 
8209
8213
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
8210
8214
  _mm_mask3_fnmsub_sd (__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
@@ -8217,36 +8221,36 @@ _mm_mask3_fnmsub_sd (__m128d __W, __m128d __X, __m128d __Y, __mmask8 __U)
8217
8221
  }
8218
8222
 
8219
8223
  #define _mm_mask3_fnmsub_round_sd(W, X, Y, U, R) \
8220
- (__m128d)__builtin_ia32_vfmsubsd3_mask3((__v2df)(__m128d)(W), \
8221
- -(__v2df)(__m128d)(X), \
8222
- (__v2df)(__m128d)(Y), \
8223
- (__mmask8)(U), (int)(R))
8224
+ ((__m128d)__builtin_ia32_vfmsubsd3_mask3((__v2df)(__m128d)(W), \
8225
+ -(__v2df)(__m128d)(X), \
8226
+ (__v2df)(__m128d)(Y), \
8227
+ (__mmask8)(U), (int)(R)))
8224
8228
 
8225
8229
  #define _mm512_permutex_pd(X, C) \
8226
- (__m512d)__builtin_ia32_permdf512((__v8df)(__m512d)(X), (int)(C))
8230
+ ((__m512d)__builtin_ia32_permdf512((__v8df)(__m512d)(X), (int)(C)))
8227
8231
 
8228
8232
  #define _mm512_mask_permutex_pd(W, U, X, C) \
8229
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
8230
- (__v8df)_mm512_permutex_pd((X), (C)), \
8231
- (__v8df)(__m512d)(W))
8233
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
8234
+ (__v8df)_mm512_permutex_pd((X), (C)), \
8235
+ (__v8df)(__m512d)(W)))
8232
8236
 
8233
8237
  #define _mm512_maskz_permutex_pd(U, X, C) \
8234
- (__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
8235
- (__v8df)_mm512_permutex_pd((X), (C)), \
8236
- (__v8df)_mm512_setzero_pd())
8238
+ ((__m512d)__builtin_ia32_selectpd_512((__mmask8)(U), \
8239
+ (__v8df)_mm512_permutex_pd((X), (C)), \
8240
+ (__v8df)_mm512_setzero_pd()))
8237
8241
 
8238
8242
  #define _mm512_permutex_epi64(X, C) \
8239
- (__m512i)__builtin_ia32_permdi512((__v8di)(__m512i)(X), (int)(C))
8243
+ ((__m512i)__builtin_ia32_permdi512((__v8di)(__m512i)(X), (int)(C)))
8240
8244
 
8241
8245
  #define _mm512_mask_permutex_epi64(W, U, X, C) \
8242
- (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
8243
- (__v8di)_mm512_permutex_epi64((X), (C)), \
8244
- (__v8di)(__m512i)(W))
8246
+ ((__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
8247
+ (__v8di)_mm512_permutex_epi64((X), (C)), \
8248
+ (__v8di)(__m512i)(W)))
8245
8249
 
8246
8250
  #define _mm512_maskz_permutex_epi64(U, X, C) \
8247
- (__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
8248
- (__v8di)_mm512_permutex_epi64((X), (C)), \
8249
- (__v8di)_mm512_setzero_si512())
8251
+ ((__m512i)__builtin_ia32_selectq_512((__mmask8)(U), \
8252
+ (__v8di)_mm512_permutex_epi64((X), (C)), \
8253
+ (__v8di)_mm512_setzero_si512()))
8250
8254
 
8251
8255
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
8252
8256
  _mm512_permutexvar_pd (__m512i __X, __m512d __Y)
@@ -8416,10 +8420,10 @@ _mm512_kxor (__mmask16 __A, __mmask16 __B)
8416
8420
  #define _kxor_mask16 _mm512_kxor
8417
8421
 
8418
8422
  #define _kshiftli_mask16(A, I) \
8419
- (__mmask16)__builtin_ia32_kshiftlihi((__mmask16)(A), (unsigned int)(I))
8423
+ ((__mmask16)__builtin_ia32_kshiftlihi((__mmask16)(A), (unsigned int)(I)))
8420
8424
 
8421
8425
  #define _kshiftri_mask16(A, I) \
8422
- (__mmask16)__builtin_ia32_kshiftrihi((__mmask16)(A), (unsigned int)(I))
8426
+ ((__mmask16)__builtin_ia32_kshiftrihi((__mmask16)(A), (unsigned int)(I)))
8423
8427
 
8424
8428
  static __inline__ unsigned int __DEFAULT_FN_ATTRS
8425
8429
  _cvtmask16_u32(__mmask16 __A) {
@@ -8538,48 +8542,48 @@ _mm512_maskz_compress_epi32 (__mmask16 __U, __m512i __A)
8538
8542
  }
8539
8543
 
8540
8544
  #define _mm_cmp_round_ss_mask(X, Y, P, R) \
8541
- (__mmask8)__builtin_ia32_cmpss_mask((__v4sf)(__m128)(X), \
8542
- (__v4sf)(__m128)(Y), (int)(P), \
8543
- (__mmask8)-1, (int)(R))
8545
+ ((__mmask8)__builtin_ia32_cmpss_mask((__v4sf)(__m128)(X), \
8546
+ (__v4sf)(__m128)(Y), (int)(P), \
8547
+ (__mmask8)-1, (int)(R)))
8544
8548
 
8545
8549
  #define _mm_mask_cmp_round_ss_mask(M, X, Y, P, R) \
8546
- (__mmask8)__builtin_ia32_cmpss_mask((__v4sf)(__m128)(X), \
8547
- (__v4sf)(__m128)(Y), (int)(P), \
8548
- (__mmask8)(M), (int)(R))
8550
+ ((__mmask8)__builtin_ia32_cmpss_mask((__v4sf)(__m128)(X), \
8551
+ (__v4sf)(__m128)(Y), (int)(P), \
8552
+ (__mmask8)(M), (int)(R)))
8549
8553
 
8550
8554
  #define _mm_cmp_ss_mask(X, Y, P) \
8551
- (__mmask8)__builtin_ia32_cmpss_mask((__v4sf)(__m128)(X), \
8552
- (__v4sf)(__m128)(Y), (int)(P), \
8553
- (__mmask8)-1, \
8554
- _MM_FROUND_CUR_DIRECTION)
8555
+ ((__mmask8)__builtin_ia32_cmpss_mask((__v4sf)(__m128)(X), \
8556
+ (__v4sf)(__m128)(Y), (int)(P), \
8557
+ (__mmask8)-1, \
8558
+ _MM_FROUND_CUR_DIRECTION))
8555
8559
 
8556
8560
  #define _mm_mask_cmp_ss_mask(M, X, Y, P) \
8557
- (__mmask8)__builtin_ia32_cmpss_mask((__v4sf)(__m128)(X), \
8558
- (__v4sf)(__m128)(Y), (int)(P), \
8559
- (__mmask8)(M), \
8560
- _MM_FROUND_CUR_DIRECTION)
8561
+ ((__mmask8)__builtin_ia32_cmpss_mask((__v4sf)(__m128)(X), \
8562
+ (__v4sf)(__m128)(Y), (int)(P), \
8563
+ (__mmask8)(M), \
8564
+ _MM_FROUND_CUR_DIRECTION))
8561
8565
 
8562
8566
  #define _mm_cmp_round_sd_mask(X, Y, P, R) \
8563
- (__mmask8)__builtin_ia32_cmpsd_mask((__v2df)(__m128d)(X), \
8564
- (__v2df)(__m128d)(Y), (int)(P), \
8565
- (__mmask8)-1, (int)(R))
8567
+ ((__mmask8)__builtin_ia32_cmpsd_mask((__v2df)(__m128d)(X), \
8568
+ (__v2df)(__m128d)(Y), (int)(P), \
8569
+ (__mmask8)-1, (int)(R)))
8566
8570
 
8567
8571
  #define _mm_mask_cmp_round_sd_mask(M, X, Y, P, R) \
8568
- (__mmask8)__builtin_ia32_cmpsd_mask((__v2df)(__m128d)(X), \
8569
- (__v2df)(__m128d)(Y), (int)(P), \
8570
- (__mmask8)(M), (int)(R))
8572
+ ((__mmask8)__builtin_ia32_cmpsd_mask((__v2df)(__m128d)(X), \
8573
+ (__v2df)(__m128d)(Y), (int)(P), \
8574
+ (__mmask8)(M), (int)(R)))
8571
8575
 
8572
8576
  #define _mm_cmp_sd_mask(X, Y, P) \
8573
- (__mmask8)__builtin_ia32_cmpsd_mask((__v2df)(__m128d)(X), \
8574
- (__v2df)(__m128d)(Y), (int)(P), \
8575
- (__mmask8)-1, \
8576
- _MM_FROUND_CUR_DIRECTION)
8577
+ ((__mmask8)__builtin_ia32_cmpsd_mask((__v2df)(__m128d)(X), \
8578
+ (__v2df)(__m128d)(Y), (int)(P), \
8579
+ (__mmask8)-1, \
8580
+ _MM_FROUND_CUR_DIRECTION))
8577
8581
 
8578
8582
  #define _mm_mask_cmp_sd_mask(M, X, Y, P) \
8579
- (__mmask8)__builtin_ia32_cmpsd_mask((__v2df)(__m128d)(X), \
8580
- (__v2df)(__m128d)(Y), (int)(P), \
8581
- (__mmask8)(M), \
8582
- _MM_FROUND_CUR_DIRECTION)
8583
+ ((__mmask8)__builtin_ia32_cmpsd_mask((__v2df)(__m128d)(X), \
8584
+ (__v2df)(__m128d)(Y), (int)(P), \
8585
+ (__mmask8)(M), \
8586
+ _MM_FROUND_CUR_DIRECTION))
8583
8587
 
8584
8588
  /* Bit Test */
8585
8589
 
@@ -8760,17 +8764,17 @@ _mm_maskz_load_sd (__mmask8 __U, const double* __A)
8760
8764
  }
8761
8765
 
8762
8766
  #define _mm512_shuffle_epi32(A, I) \
8763
- (__m512i)__builtin_ia32_pshufd512((__v16si)(__m512i)(A), (int)(I))
8767
+ ((__m512i)__builtin_ia32_pshufd512((__v16si)(__m512i)(A), (int)(I)))
8764
8768
 
8765
8769
  #define _mm512_mask_shuffle_epi32(W, U, A, I) \
8766
- (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
8767
- (__v16si)_mm512_shuffle_epi32((A), (I)), \
8768
- (__v16si)(__m512i)(W))
8770
+ ((__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
8771
+ (__v16si)_mm512_shuffle_epi32((A), (I)), \
8772
+ (__v16si)(__m512i)(W)))
8769
8773
 
8770
8774
  #define _mm512_maskz_shuffle_epi32(U, A, I) \
8771
- (__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
8772
- (__v16si)_mm512_shuffle_epi32((A), (I)), \
8773
- (__v16si)_mm512_setzero_si512())
8775
+ ((__m512i)__builtin_ia32_selectd_512((__mmask16)(U), \
8776
+ (__v16si)_mm512_shuffle_epi32((A), (I)), \
8777
+ (__v16si)_mm512_setzero_si512()))
8774
8778
 
8775
8779
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
8776
8780
  _mm512_mask_expand_pd (__m512d __W, __mmask8 __U, __m512d __A)
@@ -8901,19 +8905,19 @@ _mm512_maskz_expand_epi32 (__mmask16 __U, __m512i __A)
8901
8905
  }
8902
8906
 
8903
8907
  #define _mm512_cvt_roundps_pd(A, R) \
8904
- (__m512d)__builtin_ia32_cvtps2pd512_mask((__v8sf)(__m256)(A), \
8905
- (__v8df)_mm512_undefined_pd(), \
8906
- (__mmask8)-1, (int)(R))
8908
+ ((__m512d)__builtin_ia32_cvtps2pd512_mask((__v8sf)(__m256)(A), \
8909
+ (__v8df)_mm512_undefined_pd(), \
8910
+ (__mmask8)-1, (int)(R)))
8907
8911
 
8908
8912
  #define _mm512_mask_cvt_roundps_pd(W, U, A, R) \
8909
- (__m512d)__builtin_ia32_cvtps2pd512_mask((__v8sf)(__m256)(A), \
8910
- (__v8df)(__m512d)(W), \
8911
- (__mmask8)(U), (int)(R))
8913
+ ((__m512d)__builtin_ia32_cvtps2pd512_mask((__v8sf)(__m256)(A), \
8914
+ (__v8df)(__m512d)(W), \
8915
+ (__mmask8)(U), (int)(R)))
8912
8916
 
8913
8917
  #define _mm512_maskz_cvt_roundps_pd(U, A, R) \
8914
- (__m512d)__builtin_ia32_cvtps2pd512_mask((__v8sf)(__m256)(A), \
8915
- (__v8df)_mm512_setzero_pd(), \
8916
- (__mmask8)(U), (int)(R))
8918
+ ((__m512d)__builtin_ia32_cvtps2pd512_mask((__v8sf)(__m256)(A), \
8919
+ (__v8df)_mm512_setzero_pd(), \
8920
+ (__mmask8)(U), (int)(R)))
8917
8921
 
8918
8922
  static __inline__ __m512d __DEFAULT_FN_ATTRS512
8919
8923
  _mm512_cvtps_pd (__m256 __A)
@@ -9010,22 +9014,22 @@ _mm512_mask_compressstoreu_epi32 (void *__P, __mmask16 __U, __m512i __A)
9010
9014
  }
9011
9015
 
9012
9016
  #define _mm_cvt_roundsd_ss(A, B, R) \
9013
- (__m128)__builtin_ia32_cvtsd2ss_round_mask((__v4sf)(__m128)(A), \
9014
- (__v2df)(__m128d)(B), \
9015
- (__v4sf)_mm_undefined_ps(), \
9016
- (__mmask8)-1, (int)(R))
9017
+ ((__m128)__builtin_ia32_cvtsd2ss_round_mask((__v4sf)(__m128)(A), \
9018
+ (__v2df)(__m128d)(B), \
9019
+ (__v4sf)_mm_undefined_ps(), \
9020
+ (__mmask8)-1, (int)(R)))
9017
9021
 
9018
9022
  #define _mm_mask_cvt_roundsd_ss(W, U, A, B, R) \
9019
- (__m128)__builtin_ia32_cvtsd2ss_round_mask((__v4sf)(__m128)(A), \
9020
- (__v2df)(__m128d)(B), \
9021
- (__v4sf)(__m128)(W), \
9022
- (__mmask8)(U), (int)(R))
9023
+ ((__m128)__builtin_ia32_cvtsd2ss_round_mask((__v4sf)(__m128)(A), \
9024
+ (__v2df)(__m128d)(B), \
9025
+ (__v4sf)(__m128)(W), \
9026
+ (__mmask8)(U), (int)(R)))
9023
9027
 
9024
9028
  #define _mm_maskz_cvt_roundsd_ss(U, A, B, R) \
9025
- (__m128)__builtin_ia32_cvtsd2ss_round_mask((__v4sf)(__m128)(A), \
9026
- (__v2df)(__m128d)(B), \
9027
- (__v4sf)_mm_setzero_ps(), \
9028
- (__mmask8)(U), (int)(R))
9029
+ ((__m128)__builtin_ia32_cvtsd2ss_round_mask((__v4sf)(__m128)(A), \
9030
+ (__v2df)(__m128d)(B), \
9031
+ (__v4sf)_mm_setzero_ps(), \
9032
+ (__mmask8)(U), (int)(R)))
9029
9033
 
9030
9034
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
9031
9035
  _mm_mask_cvtsd_ss (__m128 __W, __mmask8 __U, __m128 __A, __m128d __B)
@@ -9058,47 +9062,47 @@ _mm_maskz_cvtsd_ss (__mmask8 __U, __m128 __A, __m128d __B)
9058
9062
 
9059
9063
  #ifdef __x86_64__
9060
9064
  #define _mm_cvt_roundi64_sd(A, B, R) \
9061
- (__m128d)__builtin_ia32_cvtsi2sd64((__v2df)(__m128d)(A), (long long)(B), \
9062
- (int)(R))
9065
+ ((__m128d)__builtin_ia32_cvtsi2sd64((__v2df)(__m128d)(A), (long long)(B), \
9066
+ (int)(R)))
9063
9067
 
9064
9068
  #define _mm_cvt_roundsi64_sd(A, B, R) \
9065
- (__m128d)__builtin_ia32_cvtsi2sd64((__v2df)(__m128d)(A), (long long)(B), \
9066
- (int)(R))
9069
+ ((__m128d)__builtin_ia32_cvtsi2sd64((__v2df)(__m128d)(A), (long long)(B), \
9070
+ (int)(R)))
9067
9071
  #endif
9068
9072
 
9069
9073
  #define _mm_cvt_roundsi32_ss(A, B, R) \
9070
- (__m128)__builtin_ia32_cvtsi2ss32((__v4sf)(__m128)(A), (int)(B), (int)(R))
9074
+ ((__m128)__builtin_ia32_cvtsi2ss32((__v4sf)(__m128)(A), (int)(B), (int)(R)))
9071
9075
 
9072
9076
  #define _mm_cvt_roundi32_ss(A, B, R) \
9073
- (__m128)__builtin_ia32_cvtsi2ss32((__v4sf)(__m128)(A), (int)(B), (int)(R))
9077
+ ((__m128)__builtin_ia32_cvtsi2ss32((__v4sf)(__m128)(A), (int)(B), (int)(R)))
9074
9078
 
9075
9079
  #ifdef __x86_64__
9076
9080
  #define _mm_cvt_roundsi64_ss(A, B, R) \
9077
- (__m128)__builtin_ia32_cvtsi2ss64((__v4sf)(__m128)(A), (long long)(B), \
9078
- (int)(R))
9081
+ ((__m128)__builtin_ia32_cvtsi2ss64((__v4sf)(__m128)(A), (long long)(B), \
9082
+ (int)(R)))
9079
9083
 
9080
9084
  #define _mm_cvt_roundi64_ss(A, B, R) \
9081
- (__m128)__builtin_ia32_cvtsi2ss64((__v4sf)(__m128)(A), (long long)(B), \
9082
- (int)(R))
9085
+ ((__m128)__builtin_ia32_cvtsi2ss64((__v4sf)(__m128)(A), (long long)(B), \
9086
+ (int)(R)))
9083
9087
  #endif
9084
9088
 
9085
9089
  #define _mm_cvt_roundss_sd(A, B, R) \
9086
- (__m128d)__builtin_ia32_cvtss2sd_round_mask((__v2df)(__m128d)(A), \
9087
- (__v4sf)(__m128)(B), \
9088
- (__v2df)_mm_undefined_pd(), \
9089
- (__mmask8)-1, (int)(R))
9090
+ ((__m128d)__builtin_ia32_cvtss2sd_round_mask((__v2df)(__m128d)(A), \
9091
+ (__v4sf)(__m128)(B), \
9092
+ (__v2df)_mm_undefined_pd(), \
9093
+ (__mmask8)-1, (int)(R)))
9090
9094
 
9091
9095
  #define _mm_mask_cvt_roundss_sd(W, U, A, B, R) \
9092
- (__m128d)__builtin_ia32_cvtss2sd_round_mask((__v2df)(__m128d)(A), \
9093
- (__v4sf)(__m128)(B), \
9094
- (__v2df)(__m128d)(W), \
9095
- (__mmask8)(U), (int)(R))
9096
+ ((__m128d)__builtin_ia32_cvtss2sd_round_mask((__v2df)(__m128d)(A), \
9097
+ (__v4sf)(__m128)(B), \
9098
+ (__v2df)(__m128d)(W), \
9099
+ (__mmask8)(U), (int)(R)))
9096
9100
 
9097
9101
  #define _mm_maskz_cvt_roundss_sd(U, A, B, R) \
9098
- (__m128d)__builtin_ia32_cvtss2sd_round_mask((__v2df)(__m128d)(A), \
9099
- (__v4sf)(__m128)(B), \
9100
- (__v2df)_mm_setzero_pd(), \
9101
- (__mmask8)(U), (int)(R))
9102
+ ((__m128d)__builtin_ia32_cvtss2sd_round_mask((__v2df)(__m128d)(A), \
9103
+ (__v4sf)(__m128)(B), \
9104
+ (__v2df)_mm_setzero_pd(), \
9105
+ (__mmask8)(U), (int)(R)))
9102
9106
 
9103
9107
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
9104
9108
  _mm_mask_cvtss_sd (__m128d __W, __mmask8 __U, __m128d __A, __m128 __B)
@@ -9127,8 +9131,8 @@ _mm_cvtu32_sd (__m128d __A, unsigned __B)
9127
9131
 
9128
9132
  #ifdef __x86_64__
9129
9133
  #define _mm_cvt_roundu64_sd(A, B, R) \
9130
- (__m128d)__builtin_ia32_cvtusi2sd64((__v2df)(__m128d)(A), \
9131
- (unsigned long long)(B), (int)(R))
9134
+ ((__m128d)__builtin_ia32_cvtusi2sd64((__v2df)(__m128d)(A), \
9135
+ (unsigned long long)(B), (int)(R)))
9132
9136
 
9133
9137
  static __inline__ __m128d __DEFAULT_FN_ATTRS128
9134
9138
  _mm_cvtu64_sd (__m128d __A, unsigned long long __B)
@@ -9139,8 +9143,8 @@ _mm_cvtu64_sd (__m128d __A, unsigned long long __B)
9139
9143
  #endif
9140
9144
 
9141
9145
  #define _mm_cvt_roundu32_ss(A, B, R) \
9142
- (__m128)__builtin_ia32_cvtusi2ss32((__v4sf)(__m128)(A), (unsigned int)(B), \
9143
- (int)(R))
9146
+ ((__m128)__builtin_ia32_cvtusi2ss32((__v4sf)(__m128)(A), (unsigned int)(B), \
9147
+ (int)(R)))
9144
9148
 
9145
9149
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
9146
9150
  _mm_cvtu32_ss (__m128 __A, unsigned __B)
@@ -9151,8 +9155,8 @@ _mm_cvtu32_ss (__m128 __A, unsigned __B)
9151
9155
 
9152
9156
  #ifdef __x86_64__
9153
9157
  #define _mm_cvt_roundu64_ss(A, B, R) \
9154
- (__m128)__builtin_ia32_cvtusi2ss64((__v4sf)(__m128)(A), \
9155
- (unsigned long long)(B), (int)(R))
9158
+ ((__m128)__builtin_ia32_cvtusi2ss64((__v4sf)(__m128)(A), \
9159
+ (unsigned long long)(B), (int)(R)))
9156
9160
 
9157
9161
  static __inline__ __m128 __DEFAULT_FN_ATTRS128
9158
9162
  _mm_cvtu64_ss (__m128 __A, unsigned long long __B)
@@ -9320,11 +9324,11 @@ static __inline__ long long __DEFAULT_FN_ATTRS512 _mm512_reduce_mul_epi64(__m512
9320
9324
  }
9321
9325
 
9322
9326
  static __inline__ long long __DEFAULT_FN_ATTRS512 _mm512_reduce_and_epi64(__m512i __W) {
9323
- return __builtin_ia32_reduce_and_q512(__W);
9327
+ return __builtin_reduce_and((__v8di)__W);
9324
9328
  }
9325
9329
 
9326
9330
  static __inline__ long long __DEFAULT_FN_ATTRS512 _mm512_reduce_or_epi64(__m512i __W) {
9327
- return __builtin_ia32_reduce_or_q512(__W);
9331
+ return __builtin_reduce_or((__v8di)__W);
9328
9332
  }
9329
9333
 
9330
9334
  static __inline__ long long __DEFAULT_FN_ATTRS512
@@ -9342,13 +9346,13 @@ _mm512_mask_reduce_mul_epi64(__mmask8 __M, __m512i __W) {
9342
9346
  static __inline__ long long __DEFAULT_FN_ATTRS512
9343
9347
  _mm512_mask_reduce_and_epi64(__mmask8 __M, __m512i __W) {
9344
9348
  __W = _mm512_mask_mov_epi64(_mm512_set1_epi64(~0ULL), __M, __W);
9345
- return __builtin_ia32_reduce_and_q512(__W);
9349
+ return __builtin_reduce_and((__v8di)__W);
9346
9350
  }
9347
9351
 
9348
9352
  static __inline__ long long __DEFAULT_FN_ATTRS512
9349
9353
  _mm512_mask_reduce_or_epi64(__mmask8 __M, __m512i __W) {
9350
9354
  __W = _mm512_maskz_mov_epi64(__M, __W);
9351
- return __builtin_ia32_reduce_or_q512(__W);
9355
+ return __builtin_reduce_or((__v8di)__W);
9352
9356
  }
9353
9357
 
9354
9358
  // -0.0 is used to ignore the start value since it is the neutral value of
@@ -9386,12 +9390,12 @@ _mm512_reduce_mul_epi32(__m512i __W) {
9386
9390
 
9387
9391
  static __inline__ int __DEFAULT_FN_ATTRS512
9388
9392
  _mm512_reduce_and_epi32(__m512i __W) {
9389
- return __builtin_ia32_reduce_and_d512((__v16si)__W);
9393
+ return __builtin_reduce_and((__v16si)__W);
9390
9394
  }
9391
9395
 
9392
9396
  static __inline__ int __DEFAULT_FN_ATTRS512
9393
9397
  _mm512_reduce_or_epi32(__m512i __W) {
9394
- return __builtin_ia32_reduce_or_d512((__v16si)__W);
9398
+ return __builtin_reduce_or((__v16si)__W);
9395
9399
  }
9396
9400
 
9397
9401
  static __inline__ int __DEFAULT_FN_ATTRS512
@@ -9409,13 +9413,13 @@ _mm512_mask_reduce_mul_epi32( __mmask16 __M, __m512i __W) {
9409
9413
  static __inline__ int __DEFAULT_FN_ATTRS512
9410
9414
  _mm512_mask_reduce_and_epi32( __mmask16 __M, __m512i __W) {
9411
9415
  __W = _mm512_mask_mov_epi32(_mm512_set1_epi32(~0U), __M, __W);
9412
- return __builtin_ia32_reduce_and_d512((__v16si)__W);
9416
+ return __builtin_reduce_and((__v16si)__W);
9413
9417
  }
9414
9418
 
9415
9419
  static __inline__ int __DEFAULT_FN_ATTRS512
9416
9420
  _mm512_mask_reduce_or_epi32(__mmask16 __M, __m512i __W) {
9417
9421
  __W = _mm512_maskz_mov_epi32(__M, __W);
9418
- return __builtin_ia32_reduce_or_d512((__v16si)__W);
9422
+ return __builtin_reduce_or((__v16si)__W);
9419
9423
  }
9420
9424
 
9421
9425
  static __inline__ float __DEFAULT_FN_ATTRS512
@@ -9442,89 +9446,89 @@ _mm512_mask_reduce_mul_ps(__mmask16 __M, __m512 __W) {
9442
9446
 
9443
9447
  static __inline__ long long __DEFAULT_FN_ATTRS512
9444
9448
  _mm512_reduce_max_epi64(__m512i __V) {
9445
- return __builtin_ia32_reduce_smax_q512(__V);
9449
+ return __builtin_reduce_max((__v8di)__V);
9446
9450
  }
9447
9451
 
9448
9452
  static __inline__ unsigned long long __DEFAULT_FN_ATTRS512
9449
9453
  _mm512_reduce_max_epu64(__m512i __V) {
9450
- return __builtin_ia32_reduce_umax_q512(__V);
9454
+ return __builtin_reduce_max((__v8du)__V);
9451
9455
  }
9452
9456
 
9453
9457
  static __inline__ long long __DEFAULT_FN_ATTRS512
9454
9458
  _mm512_reduce_min_epi64(__m512i __V) {
9455
- return __builtin_ia32_reduce_smin_q512(__V);
9459
+ return __builtin_reduce_min((__v8di)__V);
9456
9460
  }
9457
9461
 
9458
9462
  static __inline__ unsigned long long __DEFAULT_FN_ATTRS512
9459
9463
  _mm512_reduce_min_epu64(__m512i __V) {
9460
- return __builtin_ia32_reduce_umin_q512(__V);
9464
+ return __builtin_reduce_min((__v8du)__V);
9461
9465
  }
9462
9466
 
9463
9467
  static __inline__ long long __DEFAULT_FN_ATTRS512
9464
9468
  _mm512_mask_reduce_max_epi64(__mmask8 __M, __m512i __V) {
9465
9469
  __V = _mm512_mask_mov_epi64(_mm512_set1_epi64(-__LONG_LONG_MAX__ - 1LL), __M, __V);
9466
- return __builtin_ia32_reduce_smax_q512(__V);
9470
+ return __builtin_reduce_max((__v8di)__V);
9467
9471
  }
9468
9472
 
9469
9473
  static __inline__ unsigned long long __DEFAULT_FN_ATTRS512
9470
9474
  _mm512_mask_reduce_max_epu64(__mmask8 __M, __m512i __V) {
9471
9475
  __V = _mm512_maskz_mov_epi64(__M, __V);
9472
- return __builtin_ia32_reduce_umax_q512(__V);
9476
+ return __builtin_reduce_max((__v8du)__V);
9473
9477
  }
9474
9478
 
9475
9479
  static __inline__ long long __DEFAULT_FN_ATTRS512
9476
9480
  _mm512_mask_reduce_min_epi64(__mmask8 __M, __m512i __V) {
9477
9481
  __V = _mm512_mask_mov_epi64(_mm512_set1_epi64(__LONG_LONG_MAX__), __M, __V);
9478
- return __builtin_ia32_reduce_smin_q512(__V);
9482
+ return __builtin_reduce_min((__v8di)__V);
9479
9483
  }
9480
9484
 
9481
9485
  static __inline__ unsigned long long __DEFAULT_FN_ATTRS512
9482
9486
  _mm512_mask_reduce_min_epu64(__mmask8 __M, __m512i __V) {
9483
9487
  __V = _mm512_mask_mov_epi64(_mm512_set1_epi64(~0ULL), __M, __V);
9484
- return __builtin_ia32_reduce_umin_q512(__V);
9488
+ return __builtin_reduce_min((__v8du)__V);
9485
9489
  }
9486
9490
  static __inline__ int __DEFAULT_FN_ATTRS512
9487
9491
  _mm512_reduce_max_epi32(__m512i __V) {
9488
- return __builtin_ia32_reduce_smax_d512((__v16si)__V);
9492
+ return __builtin_reduce_max((__v16si)__V);
9489
9493
  }
9490
9494
 
9491
9495
  static __inline__ unsigned int __DEFAULT_FN_ATTRS512
9492
9496
  _mm512_reduce_max_epu32(__m512i __V) {
9493
- return __builtin_ia32_reduce_umax_d512((__v16si)__V);
9497
+ return __builtin_reduce_max((__v16su)__V);
9494
9498
  }
9495
9499
 
9496
9500
  static __inline__ int __DEFAULT_FN_ATTRS512
9497
9501
  _mm512_reduce_min_epi32(__m512i __V) {
9498
- return __builtin_ia32_reduce_smin_d512((__v16si)__V);
9502
+ return __builtin_reduce_min((__v16si)__V);
9499
9503
  }
9500
9504
 
9501
9505
  static __inline__ unsigned int __DEFAULT_FN_ATTRS512
9502
9506
  _mm512_reduce_min_epu32(__m512i __V) {
9503
- return __builtin_ia32_reduce_umin_d512((__v16si)__V);
9507
+ return __builtin_reduce_min((__v16su)__V);
9504
9508
  }
9505
9509
 
9506
9510
  static __inline__ int __DEFAULT_FN_ATTRS512
9507
9511
  _mm512_mask_reduce_max_epi32(__mmask16 __M, __m512i __V) {
9508
9512
  __V = _mm512_mask_mov_epi32(_mm512_set1_epi32(-__INT_MAX__ - 1), __M, __V);
9509
- return __builtin_ia32_reduce_smax_d512((__v16si)__V);
9513
+ return __builtin_reduce_max((__v16si)__V);
9510
9514
  }
9511
9515
 
9512
9516
  static __inline__ unsigned int __DEFAULT_FN_ATTRS512
9513
9517
  _mm512_mask_reduce_max_epu32(__mmask16 __M, __m512i __V) {
9514
9518
  __V = _mm512_maskz_mov_epi32(__M, __V);
9515
- return __builtin_ia32_reduce_umax_d512((__v16si)__V);
9519
+ return __builtin_reduce_max((__v16su)__V);
9516
9520
  }
9517
9521
 
9518
9522
  static __inline__ int __DEFAULT_FN_ATTRS512
9519
9523
  _mm512_mask_reduce_min_epi32(__mmask16 __M, __m512i __V) {
9520
9524
  __V = _mm512_mask_mov_epi32(_mm512_set1_epi32(__INT_MAX__), __M, __V);
9521
- return __builtin_ia32_reduce_smin_d512((__v16si)__V);
9525
+ return __builtin_reduce_min((__v16si)__V);
9522
9526
  }
9523
9527
 
9524
9528
  static __inline__ unsigned int __DEFAULT_FN_ATTRS512
9525
9529
  _mm512_mask_reduce_min_epu32(__mmask16 __M, __m512i __V) {
9526
9530
  __V = _mm512_mask_mov_epi32(_mm512_set1_epi32(~0U), __M, __V);
9527
- return __builtin_ia32_reduce_umin_d512((__v16si)__V);
9531
+ return __builtin_reduce_min((__v16su)__V);
9528
9532
  }
9529
9533
 
9530
9534
  static __inline__ double __DEFAULT_FN_ATTRS512