[CLANG][avx512][BUILTIN] Adding fixupimm{pd|ps|sd|ss}

getexp{sd|ss} getmant{sd|ss} kunpck{di|si} loada{pd|ps} loaddqu{di|hi|qi|si} max{sd|ss} min{sd|ss} kmov16 builtins to clang


Differential Revision: http://reviews.llvm.org/D18215

llvm-svn: 264574

4 files changed, 686 insertions, 0 deletions

diff --git a/clang/lib/Headers/avx512bwintrin.h b/clang/lib/Headers/avx512bwintrin.h
index afa79b7d173..f641c71256c 100644
--- a/clang/lib/Headers/avx512bwintrin.h
+++ b/clang/lib/Headers/avx512bwintrin.h
@@ -1934,6 +1934,53 @@ _mm512_maskz_set1_epi8 (__mmask64 __M, char __A)
                  __M);
 }
 
+static __inline__ __mmask64 __DEFAULT_FN_ATTRS
+_mm512_kunpackd (__mmask64 __A, __mmask64 __B)
+{
+  return (__mmask64) __builtin_ia32_kunpckdi ((__mmask64) __A,
+                (__mmask64) __B);
+}
+
+static __inline__ __mmask32 __DEFAULT_FN_ATTRS
+_mm512_kunpackw (__mmask32 __A, __mmask32 __B)
+{
+  return (__mmask32) __builtin_ia32_kunpcksi ((__mmask32) __A,
+                (__mmask32) __B);
+}
+
+static __inline__ __m512i __DEFAULT_FN_ATTRS
+_mm512_mask_loadu_epi16 (__m512i __W, __mmask32 __U, void const *__P)
+{
+  return (__m512i) __builtin_ia32_loaddquhi512_mask ((__v32hi *) __P,
+                 (__v32hi) __W,
+                 (__mmask32) __U);
+}
+
+static __inline__ __m512i __DEFAULT_FN_ATTRS
+_mm512_maskz_loadu_epi16 (__mmask32 __U, void const *__P)
+{
+  return (__m512i) __builtin_ia32_loaddquhi512_mask ((__v32hi *) __P,
+                 (__v32hi)
+                 _mm512_setzero_hi (),
+                 (__mmask32) __U);
+}
+
+static __inline__ __m512i __DEFAULT_FN_ATTRS
+_mm512_mask_loadu_epi8 (__m512i __W, __mmask64 __U, void const *__P)
+{
+  return (__m512i) __builtin_ia32_loaddquqi512_mask ((__v64qi *) __P,
+                 (__v64qi) __W,
+                 (__mmask64) __U);
+}
+
+static __inline__ __m512i __DEFAULT_FN_ATTRS
+_mm512_maskz_loadu_epi8 (__mmask64 __U, void const *__P)
+{
+  return (__m512i) __builtin_ia32_loaddquqi512_mask ((__v64qi *) __P,
+                 (__v64qi)
+                 _mm512_setzero_hi (),
+                 (__mmask64) __U);
+}
 #undef __DEFAULT_FN_ATTRS
 
 #endif
diff --git a/clang/lib/Headers/avx512fintrin.h b/clang/lib/Headers/avx512fintrin.h
index 22a27299e24..e1f81a0e2a1 100644
--- a/clang/lib/Headers/avx512fintrin.h
+++ b/clang/lib/Headers/avx512fintrin.h
@@ -46,6 +46,21 @@ typedef unsigned short __mmask16;
 #define _MM_FROUND_TO_ZERO          0x03
 #define _MM_FROUND_CUR_DIRECTION    0x04
 
+typedef enum
+{
+  _MM_MANT_NORM_1_2,    /* interval [1, 2)      */
+  _MM_MANT_NORM_p5_2,   /* interval [0.5, 2)    */
+  _MM_MANT_NORM_p5_1,   /* interval [0.5, 1)    */
+  _MM_MANT_NORM_p75_1p5   /* interval [0.75, 1.5) */
+} _MM_MANTISSA_NORM_ENUM;
+
+typedef enum
+{
+  _MM_MANT_SIGN_src,    /* sign = sign(SRC)     */
+  _MM_MANT_SIGN_zero,   /* sign = 0             */
+  _MM_MANT_SIGN_nan   /* DEST = NaN if sign(SRC) = 1 */
+} _MM_MANTISSA_SIGN_ENUM;
+
 /* Define the default attributes for the functions in this file. */
 #define __DEFAULT_FN_ATTRS __attribute__((__always_inline__, __nodebug__, __target__("avx512f")))
 
@@ -3695,6 +3710,262 @@ _mm512_maskz_movedup_pd (__mmask8 __U, __m512d __A)
                (__mmask8) __U);
 }
 
+#define _mm512_fixupimm_round_pd( __A, __B, __C, __imm, __R) __extension__ ({ \
+__builtin_ia32_fixupimmpd512_mask ((__v8df)( __A),\
+                  (__v8df)( __B),\
+                  (__v8di)( __C),\
+                  (__imm),\
+                  (__mmask8) -1, (__R));\
+})
+
+#define _mm512_mask_fixupimm_round_pd( __A, __U, __B, __C, __imm, __R) __extension__ ({ \
+__builtin_ia32_fixupimmpd512_mask ((__v8df)( __A),\
+                  (__v8df)( __B),\
+                  (__v8di)( __C),\
+                  (__imm),\
+                  (__mmask8)( __U), (__R));\
+})
+
+#define _mm512_fixupimm_pd( __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmpd512_mask ((__v8df)( __A),\
+                  (__v8df)( __B),\
+                  (__v8di)( __C),\
+                 ( __imm),\
+                  (__mmask8) -1,\
+                  _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm512_mask_fixupimm_pd( __A, __U, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmpd512_mask ((__v8df)( __A),\
+                  (__v8df)( __B),\
+                  (__v8di)( __C),\
+                 ( __imm),\
+                  (__mmask8)( __U),\
+                  _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm512_maskz_fixupimm_round_pd( __U, __A, __B, __C, __imm, __R) __extension__ ({ \
+__builtin_ia32_fixupimmpd512_maskz ((__v8df)( __A),\
+                   (__v8df)( __B),\
+                   (__v8di)( __C),\
+                   (__imm),\
+                   (__mmask8)( __U), (__R));\
+})
+
+#define _mm512_maskz_fixupimm_pd( __U, __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmpd512_maskz ((__v8df)( __A),\
+                   (__v8df)( __B),\
+                   (__v8di)( __C),\
+                  ( __imm),\
+                   (__mmask8)( __U),\
+                   _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm512_fixupimm_round_ps( __A, __B, __C, __imm, __R) __extension__ ({ \
+__builtin_ia32_fixupimmps512_mask ((__v16sf)( __A),\
+                 (__v16sf)( __B),\
+                 (__v16si)( __C),\
+                 (__imm),\
+                 (__mmask16) -1, (__R));\
+})
+
+#define _mm512_mask_fixupimm_round_ps( __A, __U, __B, __C, __imm, __R) __extension__ ({ \
+__builtin_ia32_fixupimmps512_mask ((__v16sf)( __A),\
+                 (__v16sf)( __B),\
+                 (__v16si)( __C),\
+                 (__imm),\
+                 (__mmask16)( __U), (__R));\
+})
+
+#define _mm512_fixupimm_ps( __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmps512_mask ((__v16sf)( __A),\
+                 (__v16sf)( __B),\
+                 (__v16si)( __C),\
+                ( __imm),\
+                 (__mmask16) -1,\
+                 _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm512_mask_fixupimm_ps( __A, __U, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmps512_mask ((__v16sf)( __A),\
+                 (__v16sf)( __B),\
+                 (__v16si)( __C),\
+                ( __imm),\
+                 (__mmask16)( __U),\
+                 _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm512_maskz_fixupimm_round_ps( __U, __A, __B, __C, __imm, __R) __extension__ ({ \
+__builtin_ia32_fixupimmps512_maskz ((__v16sf)( __A),\
+                  (__v16sf)( __B),\
+                  (__v16si)( __C),\
+                  (__imm),\
+                  (__mmask16)( __U), (__R));\
+})
+
+#define _mm512_maskz_fixupimm_ps( __U, __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmps512_maskz ((__v16sf)( __A),\
+                  (__v16sf)( __B),\
+                  (__v16si)( __C),\
+                 ( __imm),\
+                  (__mmask16)( __U),\
+                  _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm_fixupimm_round_sd( __A, __B, __C, __imm, __R) __extension__ ({ \
+__builtin_ia32_fixupimmsd_mask ((__v2df)( __A),\
+               (__v2df)( __B),\
+               (__v2di)( __C), __imm,\
+               (__mmask8) -1, (__R));\
+})
+
+#define _mm_mask_fixupimm_round_sd( __A, __U, __B, __C, __imm, __R) __extension__ ({ \
+__builtin_ia32_fixupimmsd_mask ((__v2df)( __A),\
+               (__v2df)( __B),\
+               (__v2di)( __C), __imm,\
+               (__mmask8)( __U), (__R));\
+})
+
+#define _mm_fixupimm_sd( __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmsd_mask ((__v2df)( __A),\
+               (__v2df)( __B),\
+               (__v2di)( __C),( __imm),\
+               (__mmask8) -1,\
+               _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm_mask_fixupimm_sd( __A, __U, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmsd_mask ((__v2df)( __A),\
+               (__v2df)( __B),\
+               (__v2di)( __C),( __imm),\
+               (__mmask8)( __U),\
+               _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm_maskz_fixupimm_round_sd( __U, __A, __B, __C, __imm, __R) __extension__ ({ \
+__builtin_ia32_fixupimmsd_maskz ((__v2df)( __A),\
+                (__v2df)( __B),\
+                (__v2di)( __C),\
+                __imm,\
+                (__mmask8)( __U), (__R));\
+})
+
+#define _mm_maskz_fixupimm_sd( __U, __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmsd_maskz ((__v2df)( __A),\
+                (__v2df)( __B),\
+                (__v2di)( __C),\
+               ( __imm),\
+                (__mmask8)( __U),\
+                _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm_fixupimm_round_ss( __A, __B, __C, __imm, __R) __extension__ ({ \
+__builtin_ia32_fixupimmss_mask ((__v4sf)( __A),\
+              (__v4sf)( __B),\
+              (__v4si)( __C), (__imm),\
+              (__mmask8) -1, (__R));\
+})
+
+#define _mm_mask_fixupimm_round_ss( __A, __U, __B, __C, __imm, __R) __extension__ ({ \
+__builtin_ia32_fixupimmss_mask ((__v4sf)( __A),\
+              (__v4sf)( __B),\
+              (__v4si)( __C), (__imm),\
+              (__mmask8)( __U), (__R));\
+})
+
+#define _mm_fixupimm_ss( __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmss_mask ((__v4sf)( __A),\
+              (__v4sf)( __B),\
+              (__v4si)( __C),( __imm),\
+              (__mmask8) -1,\
+              _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm_mask_fixupimm_ss( __A, __U, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmss_mask ((__v4sf)( __A),\
+              (__v4sf)( __B),\
+              (__v4si)( __C),( __imm),\
+              (__mmask8)( __U),\
+              _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm_maskz_fixupimm_round_ss( __U, __A, __B, __C, __imm, __R) __extension__ ({ \
+__builtin_ia32_fixupimmss_maskz ((__v4sf)( __A),\
+               (__v4sf)( __B),\
+               (__v4si)( __C), (__imm),\
+               (__mmask8)( __U), (__R));\
+})
+
+#define _mm_maskz_fixupimm_ss( __U, __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmss_maskz ((__v4sf)( __A),\
+               (__v4sf)( __B),\
+               (__v4si)( __C),( __imm),\
+               (__mmask8)( __U),\
+               _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm_getexp_round_sd( __A, __B ,__R) __extension__ ({ \
+__builtin_ia32_getexpsd128_round_mask ((__v2df)(__A),\
+                 (__v2df)( __B), (__v2df) _mm_setzero_pd(), (__mmask8) -1,\
+                 ( __R));\
+})
+
+
+static __inline__ __m128d __DEFAULT_FN_ATTRS
+_mm_getexp_sd (__m128d __A, __m128d __B)
+{
+  return (__m128d) __builtin_ia32_getexpsd128_round_mask ((__v2df) __A,
+                 (__v2df) __B, (__v2df) _mm_setzero_pd(), (__mmask8) -1, _MM_FROUND_CUR_DIRECTION);
+}
+
+#define _mm_getexp_round_ss( __A, __B, __R) __extension__ ({ \
+__builtin_ia32_getexpss128_round_mask ((__v4sf)( __A),\
+                (__v4sf)( __B), (__v4sf)  _mm_setzero_ps(), (__mmask8) -1,\
+                ( __R));\
+})
+
+static __inline__ __m128 __DEFAULT_FN_ATTRS
+_mm_getexp_ss (__m128 __A, __m128 __B)
+{
+  return (__m128) __builtin_ia32_getexpss128_round_mask ((__v4sf) __A,
+                (__v4sf) __B, (__v4sf)  _mm_setzero_ps(), (__mmask8) -1, _MM_FROUND_CUR_DIRECTION);
+}
+
+#define _mm_getmant_round_sd( __A, __B, __C, __D, __R) __extension__ ({ \
+__builtin_ia32_getmantsd_round_mask ((__v2df)( __A),\
+              (__v2df)( __B),\
+              (( __D) << 2) |( __C), (__v2df) _mm_setzero_pd(), (__mmask8) -1,\
+              ( __R));\
+})
+
+#define _mm_getmant_sd( __A,  __B,  __C, __D)  __extension__ ({ \
+__builtin_ia32_getmantsd_round_mask ((__v2df)( __A),\
+              (__v2df)( __B),\
+              (( __D) << 2) |( __C), (__v2df) _mm_setzero_pd(), (__mmask8) -1,\
+              _MM_FROUND_CUR_DIRECTION);\
+})
+
+#define _mm_getmant_round_ss( __A, __B, __C, __D, __R) __extension__ ({ \
+__builtin_ia32_getmantss_round_mask ((__v4sf)( __A),\
+              (__v4sf)( __B),\
+              ((__D) << 2) |( __C), (__v4sf)  _mm_setzero_ps(), (__mmask8) -1,\
+              ( __R));\
+})
+
+#define  _mm_getmant_ss(__A,  __B,  __C, __D) __extension__ ({ \
+__builtin_ia32_getmantss_round_mask ((__v4sf)( __A),\
+              (__v4sf)( __B),\
+              ((__D) << 2) |( __C), (__v4sf)  _mm_setzero_ps(), (__mmask8) -1,\
+              _MM_FROUND_CUR_DIRECTION);\
+})
+
+
+static __inline__ __mmask16 __DEFAULT_FN_ATTRS
+_mm512_kmov (__mmask16 __A)
+{
+  return  __A;
+}
+
 #define _mm_comi_round_sd(__A, __B, __P, __R) __extension__ ({\
 __builtin_ia32_vcomisd ((__v2df) (__A), (__v2df) (__B), ( __P), (  __R));\
 })
diff --git a/clang/lib/Headers/avx512vlbwintrin.h b/clang/lib/Headers/avx512vlbwintrin.h
index 9209de38cca..bb69b1bfdd3 100644
--- a/clang/lib/Headers/avx512vlbwintrin.h
+++ b/clang/lib/Headers/avx512vlbwintrin.h
@@ -2952,6 +2952,74 @@ _mm256_maskz_set1_epi8 (__mmask32 __M, char __A)
                  _mm256_setzero_si256 (),
                  __M);
 }
+
+static __inline__ __m128i __DEFAULT_FN_ATTRS
+_mm_mask_loadu_epi16 (__m128i __W, __mmask8 __U, void const *__P)
+{
+  return (__m128i) __builtin_ia32_loaddquhi128_mask ((__v8hi *) __P,
+                 (__v8hi) __W,
+                 (__mmask8) __U);
+}
+
+static __inline__ __m128i __DEFAULT_FN_ATTRS
+_mm_maskz_loadu_epi16 (__mmask8 __U, void const *__P)
+{
+  return (__m128i) __builtin_ia32_loaddquhi128_mask ((__v8hi *) __P,
+                 (__v8hi)
+                 _mm_setzero_hi (),
+                 (__mmask8) __U);
+}
+
+static __inline__ __m256i __DEFAULT_FN_ATTRS
+_mm256_mask_loadu_epi16 (__m256i __W, __mmask16 __U, void const *__P)
+{
+  return (__m256i) __builtin_ia32_loaddquhi256_mask ((__v16hi *) __P,
+                 (__v16hi) __W,
+                 (__mmask16) __U);
+}
+
+static __inline__ __m256i __DEFAULT_FN_ATTRS
+_mm256_maskz_loadu_epi16 (__mmask16 __U, void const *__P)
+{
+  return (__m256i) __builtin_ia32_loaddquhi256_mask ((__v16hi *) __P,
+                 (__v16hi)
+                 _mm256_setzero_si256 (),
+                 (__mmask16) __U);
+}
+
+static __inline__ __m128i __DEFAULT_FN_ATTRS
+_mm_mask_loadu_epi8 (__m128i __W, __mmask16 __U, void const *__P)
+{
+  return (__m128i) __builtin_ia32_loaddquqi128_mask ((__v16qi *) __P,
+                 (__v16qi) __W,
+                 (__mmask16) __U);
+}
+
+static __inline__ __m128i __DEFAULT_FN_ATTRS
+_mm_maskz_loadu_epi8 (__mmask16 __U, void const *__P)
+{
+  return (__m128i) __builtin_ia32_loaddquqi128_mask ((__v16qi *) __P,
+                 (__v16qi)
+                 _mm_setzero_si128 (),
+                 (__mmask16) __U);
+}
+
+static __inline__ __m256i __DEFAULT_FN_ATTRS
+_mm256_mask_loadu_epi8 (__m256i __W, __mmask32 __U, void const *__P)
+{
+  return (__m256i) __builtin_ia32_loaddquqi256_mask ((__v32qi *) __P,
+                 (__v32qi) __W,
+                 (__mmask32) __U);
+}
+
+static __inline__ __m256i __DEFAULT_FN_ATTRS
+_mm256_maskz_loadu_epi8 (__mmask32 __U, void const *__P)
+{
+  return (__m256i) __builtin_ia32_loaddquqi256_mask ((__v32qi *) __P,
+                 (__v32qi)
+                 _mm256_setzero_si256 (),
+                 (__mmask32) __U);
+}
 #undef __DEFAULT_FN_ATTRS
 
 #endif /* __AVX512VLBWINTRIN_H */
diff --git a/clang/lib/Headers/avx512vlintrin.h b/clang/lib/Headers/avx512vlintrin.h
index 0949cb9238a..6080a6d5324 100644
--- a/clang/lib/Headers/avx512vlintrin.h
+++ b/clang/lib/Headers/avx512vlintrin.h
@@ -6068,6 +6068,306 @@ _mm256_maskz_set1_epi64 (__mmask8 __M, long long __A)
                  _mm256_setzero_si256 (),
                  __M);
 }
+
+#define _mm_fixupimm_pd( __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmpd128_mask ((__v2df)( __A),\
+                  (__v2df)( __B),\
+                  (__v2di)( __C),\
+                 ( __imm),\
+                  (__mmask8) -1);\
+})
+
+#define _mm_mask_fixupimm_pd( __A, __U, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmpd128_mask ((__v2df)( __A),\
+                  (__v2df)( __B),\
+                  (__v2di)( __C),\
+                 ( __imm),\
+                  (__mmask8)( __U));\
+})
+
+#define _mm_maskz_fixupimm_pd( __U, __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmpd128_maskz ((__v2df)( __A),\
+                   (__v2df)( __B),\
+                   (__v2di)( __C),\
+                  ( __imm),\
+                   (__mmask8)( __U));\
+})
+
+#define _mm256_fixupimm_pd( __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmpd256_mask ((__v4df)( __A),\
+                  (__v4df)( __B),\
+                  (__v4di)( __C),\
+                 ( __imm),\
+                  (__mmask8) -1);\
+})
+
+#define _mm256_mask_fixupimm_pd( __A, __U, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmpd256_mask ((__v4df)( __A),\
+                  (__v4df)( __B),\
+                  (__v4di)( __C),\
+                 ( __imm),\
+                  (__mmask8)( __U));\
+})
+
+#define _mm256_maskz_fixupimm_pd( __U, __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmpd256_maskz ((__v4df)( __A),\
+                   (__v4df)( __B),\
+                   (__v4di)( __C),\
+                  ( __imm),\
+                   (__mmask8)( __U));\
+})
+
+#define _mm_fixupimm_ps( __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmps128_mask ((__v4sf)( __A),\
+                 (__v4sf)( __B),\
+                 (__v4si)( __C),\
+                ( __imm),\
+                 (__mmask8) -1);\
+})
+
+#define _mm_mask_fixupimm_ps( __A, __U, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmps128_mask ((__v4sf)( __A),\
+                 (__v4sf)( __B),\
+                 (__v4si)( __C),\
+                ( __imm),\
+                 (__mmask8)( __U));\
+})
+
+#define _mm_maskz_fixupimm_ps( __U, __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmps128_maskz ((__v4sf)( __A),\
+                  (__v4sf)( __B),\
+                  (__v4si)( __C),\
+                 ( __imm),\
+                  (__mmask8)( __U));\
+})
+
+#define _mm256_fixupimm_ps( __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmps256_mask ((__v8sf)( __A),\
+                 (__v8sf)( __B),\
+                 (__v8si)( __C),\
+                ( __imm),\
+                 (__mmask8) -1);\
+})
+
+#define _mm256_mask_fixupimm_ps( __A, __U, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmps256_mask ((__v8sf)( __A),\
+                 (__v8sf)( __B),\
+                 (__v8si)( __C),\
+                ( __imm),\
+                 (__mmask8)( __U));\
+})
+
+#define _mm256_maskz_fixupimm_ps( __U, __A, __B, __C, __imm) __extension__ ({ \
+__builtin_ia32_fixupimmps256_maskz ((__v8sf)( __A),\
+                  (__v8sf)( __B),\
+                  (__v8si)( __C),\
+                 ( __imm),\
+                  (__mmask8)( __U));\
+})
+
+static __inline__ __m128d __DEFAULT_FN_ATTRS
+_mm_mask_load_pd (__m128d __W, __mmask8 __U, void const *__P)
+{
+  return (__m128d) __builtin_ia32_loadapd128_mask ((__v2df *) __P,
+               (__v2df) __W,
+               (__mmask8) __U);
+}
+
+static __inline__ __m128d __DEFAULT_FN_ATTRS
+_mm_maskz_load_pd (__mmask8 __U, void const *__P)
+{
+  return (__m128d) __builtin_ia32_loadapd128_mask ((__v2df *) __P,
+               (__v2df)
+               _mm_setzero_pd (),
+               (__mmask8) __U);
+}
+
+static __inline__ __m256d __DEFAULT_FN_ATTRS
+_mm256_mask_load_pd (__m256d __W, __mmask8 __U, void const *__P)
+{
+  return (__m256d) __builtin_ia32_loadapd256_mask ((__v4df *) __P,
+               (__v4df) __W,
+               (__mmask8) __U);
+}
+
+static __inline__ __m256d __DEFAULT_FN_ATTRS
+_mm256_maskz_load_pd (__mmask8 __U, void const *__P)
+{
+  return (__m256d) __builtin_ia32_loadapd256_mask ((__v4df *) __P,
+               (__v4df)
+               _mm256_setzero_pd (),
+               (__mmask8) __U);
+}
+
+static __inline__ __m128 __DEFAULT_FN_ATTRS
+_mm_mask_load_ps (__m128 __W, __mmask8 __U, void const *__P)
+{
+  return (__m128) __builtin_ia32_loadaps128_mask ((__v4sf *) __P,
+              (__v4sf) __W,
+              (__mmask8) __U);
+}
+
+static __inline__ __m128 __DEFAULT_FN_ATTRS
+_mm_maskz_load_ps (__mmask8 __U, void const *__P)
+{
+  return (__m128) __builtin_ia32_loadaps128_mask ((__v4sf *) __P,
+              (__v4sf)
+              _mm_setzero_ps (),
+              (__mmask8) __U);
+}
+
+static __inline__ __m256 __DEFAULT_FN_ATTRS
+_mm256_mask_load_ps (__m256 __W, __mmask8 __U, void const *__P)
+{
+  return (__m256) __builtin_ia32_loadaps256_mask ((__v8sf *) __P,
+              (__v8sf) __W,
+              (__mmask8) __U);
+}
+
+static __inline__ __m256 __DEFAULT_FN_ATTRS
+_mm256_maskz_load_ps (__mmask8 __U, void const *__P)
+{
+  return (__m256) __builtin_ia32_loadaps256_mask ((__v8sf *) __P,
+              (__v8sf)
+              _mm256_setzero_ps (),
+              (__mmask8) __U);
+}
+
+static __inline__ __m128i __DEFAULT_FN_ATTRS
+_mm_mask_loadu_epi64 (__m128i __W, __mmask8 __U, void const *__P)
+{
+  return (__m128i) __builtin_ia32_loaddqudi128_mask ((__v2di *) __P,
+                 (__v2di) __W,
+                 (__mmask8) __U);
+}
+
+static __inline__ __m128i __DEFAULT_FN_ATTRS
+_mm_maskz_loadu_epi64 (__mmask8 __U, void const *__P)
+{
+  return (__m128i) __builtin_ia32_loaddqudi128_mask ((__v2di *) __P,
+                 (__v2di)
+                 _mm_setzero_si128 (),
+                 (__mmask8) __U);
+}
+
+static __inline__ __m256i __DEFAULT_FN_ATTRS
+_mm256_mask_loadu_epi64 (__m256i __W, __mmask8 __U, void const *__P)
+{
+  return (__m256i) __builtin_ia32_loaddqudi256_mask ((__v4di *) __P,
+                 (__v4di) __W,
+                 (__mmask8) __U);
+}
+
+static __inline__ __m256i __DEFAULT_FN_ATTRS
+_mm256_maskz_loadu_epi64 (__mmask8 __U, void const *__P)
+{
+  return (__m256i) __builtin_ia32_loaddqudi256_mask ((__v4di *) __P,
+                 (__v4di)
+                 _mm256_setzero_si256 (),
+                 (__mmask8) __U);
+}
+
+static __inline__ __m128i __DEFAULT_FN_ATTRS
+_mm_mask_loadu_epi32 (__m128i __W, __mmask8 __U, void const *__P)
+{
+  return (__m128i) __builtin_ia32_loaddqusi128_mask ((__v4si *) __P,
+                 (__v4si) __W,
+                 (__mmask8) __U);
+}
+
+static __inline__ __m128i __DEFAULT_FN_ATTRS
+_mm_maskz_loadu_epi32 (__mmask8 __U, void const *__P)
+{
+  return (__m128i) __builtin_ia32_loaddqusi128_mask ((__v4si *) __P,
+                 (__v4si)
+                 _mm_setzero_si128 (),
+                 (__mmask8) __U);
+}
+
+static __inline__ __m256i __DEFAULT_FN_ATTRS
+_mm256_mask_loadu_epi32 (__m256i __W, __mmask8 __U, void const *__P)
+{
+  return (__m256i) __builtin_ia32_loaddqusi256_mask ((__v8si *) __P,
+                 (__v8si) __W,
+                 (__mmask8) __U);
+}
+
+static __inline__ __m256i __DEFAULT_FN_ATTRS
+_mm256_maskz_loadu_epi32 (__mmask8 __U, void const *__P)
+{
+  return (__m256i) __builtin_ia32_loaddqusi256_mask ((__v8si *) __P,
+                 (__v8si)
+                 _mm256_setzero_si256 (),
+                 (__mmask8) __U);
+}
+
+static __inline__ __m128d __DEFAULT_FN_ATTRS
+_mm_mask_loadu_pd (__m128d __W, __mmask8 __U, void const *__P)
+{
+  return (__m128d) __builtin_ia32_loadupd128_mask ((__v2df *) __P,
+               (__v2df) __W,
+               (__mmask8) __U);
+}
+
+static __inline__ __m128d __DEFAULT_FN_ATTRS
+_mm_maskz_loadu_pd (__mmask8 __U, void const *__P)
+{
+  return (__m128d) __builtin_ia32_loadupd128_mask ((__v2df *) __P,
+               (__v2df)
+               _mm_setzero_pd (),
+               (__mmask8) __U);
+}
+
+static __inline__ __m256d __DEFAULT_FN_ATTRS
+_mm256_mask_loadu_pd (__m256d __W, __mmask8 __U, void const *__P)
+{
+  return (__m256d) __builtin_ia32_loadupd256_mask ((__v4df *) __P,
+               (__v4df) __W,
+               (__mmask8) __U);
+}
+
+static __inline__ __m256d __DEFAULT_FN_ATTRS
+_mm256_maskz_loadu_pd (__mmask8 __U, void const *__P)
+{
+  return (__m256d) __builtin_ia32_loadupd256_mask ((__v4df *) __P,
+               (__v4df)
+               _mm256_setzero_pd (),
+               (__mmask8) __U);
+}
+
+static __inline__ __m128 __DEFAULT_FN_ATTRS
+_mm_mask_loadu_ps (__m128 __W, __mmask8 __U, void const *__P)
+{
+  return (__m128) __builtin_ia32_loadups128_mask ((__v4sf *) __P,
+              (__v4sf) __W,
+              (__mmask8) __U);
+}
+
+static __inline__ __m128 __DEFAULT_FN_ATTRS
+_mm_maskz_loadu_ps (__mmask8 __U, void const *__P)
+{
+  return (__m128) __builtin_ia32_loadups128_mask ((__v4sf *) __P,
+              (__v4sf)
+              _mm_setzero_ps (),
+              (__mmask8) __U);
+}
+
+static __inline__ __m256 __DEFAULT_FN_ATTRS
+_mm256_mask_loadu_ps (__m256 __W, __mmask8 __U, void const *__P)
+{
+  return (__m256) __builtin_ia32_loadups256_mask ((__v8sf *) __P,
+              (__v8sf) __W,
+              (__mmask8) __U);
+}
+
+static __inline__ __m256 __DEFAULT_FN_ATTRS
+_mm256_maskz_loadu_ps (__mmask8 __U, void const *__P)
+{
+  return (__m256) __builtin_ia32_loadups256_mask ((__v8sf *) __P,
+              (__v8sf)
+              _mm256_setzero_ps (),
+              (__mmask8) __U);
+}
 #undef __DEFAULT_FN_ATTRS
 #undef __DEFAULT_FN_ATTRS_BOTH