@cpp-linter/clang-tidy-win32-x64 20.1.0 → 22.1.8

package/lib/clang/{20 → 22}/include/fmaintrin.h

@@ -15,8 +15,20 @@
 #define __FMAINTRIN_H
 
 /* Define the default attributes for the functions in this file. */
-#define __DEFAULT_FN_ATTRS128 __attribute__((__always_inline__, __nodebug__, __target__("fma"), __min_vector_width__(128)))
-#define __DEFAULT_FN_ATTRS256 __attribute__((__always_inline__, __nodebug__, __target__("fma"), __min_vector_width__(256)))
+#define __DEFAULT_FN_ATTRS128                                                  \
+  __attribute__((__always_inline__, __nodebug__, __target__("fma"),            \
+                 __min_vector_width__(128)))
+#define __DEFAULT_FN_ATTRS256                                                  \
+  __attribute__((__always_inline__, __nodebug__, __target__("fma"),            \
+                 __min_vector_width__(256)))
+
+#if defined(__cplusplus) && (__cplusplus >= 201103L)
+#define __DEFAULT_FN_ATTRS128_CONSTEXPR __DEFAULT_FN_ATTRS128 constexpr
+#define __DEFAULT_FN_ATTRS256_CONSTEXPR __DEFAULT_FN_ATTRS256 constexpr
+#else
+#define __DEFAULT_FN_ATTRS128_CONSTEXPR __DEFAULT_FN_ATTRS128
+#define __DEFAULT_FN_ATTRS256_CONSTEXPR __DEFAULT_FN_ATTRS256
+#endif
 
 /// Computes a multiply-add of 128-bit vectors of [4 x float].
 ///    For each element, computes <c> (__A * __B) + __C </c>.
@@ -32,10 +44,11 @@
 /// \param __C
 ///    A 128-bit vector of [4 x float] containing the addend.
 /// \returns A 128-bit vector of [4 x float] containing the result.
-static __inline__ __m128 __DEFAULT_FN_ATTRS128
+static __inline__ __m128 __DEFAULT_FN_ATTRS128_CONSTEXPR
 _mm_fmadd_ps(__m128 __A, __m128 __B, __m128 __C)
 {
-  return (__m128)__builtin_ia32_vfmaddps((__v4sf)__A, (__v4sf)__B, (__v4sf)__C);
+  return (__m128)__builtin_elementwise_fma((__v4sf)__A, (__v4sf)__B,
+                                           (__v4sf)__C);
 }
 
 /// Computes a multiply-add of 128-bit vectors of [2 x double].
@@ -52,10 +65,11 @@ _mm_fmadd_ps(__m128 __A, __m128 __B, __m128 __C)
 /// \param __C
 ///    A 128-bit vector of [2 x double] containing the addend.
 /// \returns A 128-bit [2 x double] vector containing the result.
-static __inline__ __m128d __DEFAULT_FN_ATTRS128
+static __inline__ __m128d __DEFAULT_FN_ATTRS128_CONSTEXPR
 _mm_fmadd_pd(__m128d __A, __m128d __B, __m128d __C)
 {
-  return (__m128d)__builtin_ia32_vfmaddpd((__v2df)__A, (__v2df)__B, (__v2df)__C);
+  return (__m128d)__builtin_elementwise_fma((__v2df)__A, (__v2df)__B,
+                                            (__v2df)__C);
 }
 
 /// Computes a scalar multiply-add of the single-precision values in the
@@ -81,10 +95,10 @@ _mm_fmadd_pd(__m128d __A, __m128d __B, __m128d __C)
 ///    32 bits.
 /// \returns A 128-bit vector of [4 x float] containing the result in the low
 ///    32 bits and a copy of \a __A[127:32] in the upper 96 bits.
-static __inline__ __m128 __DEFAULT_FN_ATTRS128
-_mm_fmadd_ss(__m128 __A, __m128 __B, __m128 __C)
-{
-  return (__m128)__builtin_ia32_vfmaddss3((__v4sf)__A, (__v4sf)__B, (__v4sf)__C);
+static __inline__ __m128 __DEFAULT_FN_ATTRS128_CONSTEXPR
+_mm_fmadd_ss(__m128 __A, __m128 __B, __m128 __C) {
+  __A[0] = __builtin_elementwise_fma(__A[0], __B[0], __C[0]);
+  return __A;
 }
 
 /// Computes a scalar multiply-add of the double-precision values in the
@@ -110,10 +124,10 @@ _mm_fmadd_ss(__m128 __A, __m128 __B, __m128 __C)
 ///    64 bits.
 /// \returns A 128-bit vector of [2 x double] containing the result in the low
 ///    64 bits and a copy of \a __A[127:64] in the upper 64 bits.
-static __inline__ __m128d __DEFAULT_FN_ATTRS128
-_mm_fmadd_sd(__m128d __A, __m128d __B, __m128d __C)
-{
-  return (__m128d)__builtin_ia32_vfmaddsd3((__v2df)__A, (__v2df)__B, (__v2df)__C);
+static __inline__ __m128d __DEFAULT_FN_ATTRS128_CONSTEXPR
+_mm_fmadd_sd(__m128d __A, __m128d __B, __m128d __C) {
+  __A[0] = __builtin_elementwise_fma(__A[0], __B[0], __C[0]);
+  return __A;
 }
 
 /// Computes a multiply-subtract of 128-bit vectors of [4 x float].
@@ -130,10 +144,11 @@ _mm_fmadd_sd(__m128d __A, __m128d __B, __m128d __C)
 /// \param __C
 ///    A 128-bit vector of [4 x float] containing the subtrahend.
 /// \returns A 128-bit vector of [4 x float] containing the result.
-static __inline__ __m128 __DEFAULT_FN_ATTRS128
+static __inline__ __m128 __DEFAULT_FN_ATTRS128_CONSTEXPR
 _mm_fmsub_ps(__m128 __A, __m128 __B, __m128 __C)
 {
-  return (__m128)__builtin_ia32_vfmaddps((__v4sf)__A, (__v4sf)__B, -(__v4sf)__C);
+  return (__m128)__builtin_elementwise_fma((__v4sf)__A, (__v4sf)__B,
+                                           -(__v4sf)__C);
 }
 
 /// Computes a multiply-subtract of 128-bit vectors of [2 x double].
@@ -150,10 +165,11 @@ _mm_fmsub_ps(__m128 __A, __m128 __B, __m128 __C)
 /// \param __C
 ///    A 128-bit vector of [2 x double] containing the addend.
 /// \returns A 128-bit vector of [2 x double] containing the result.
-static __inline__ __m128d __DEFAULT_FN_ATTRS128
+static __inline__ __m128d __DEFAULT_FN_ATTRS128_CONSTEXPR
 _mm_fmsub_pd(__m128d __A, __m128d __B, __m128d __C)
 {
-  return (__m128d)__builtin_ia32_vfmaddpd((__v2df)__A, (__v2df)__B, -(__v2df)__C);
+  return (__m128d)__builtin_elementwise_fma((__v2df)__A, (__v2df)__B,
+                                            -(__v2df)__C);
 }
 
 /// Computes a scalar multiply-subtract of the single-precision values in
@@ -179,10 +195,10 @@ _mm_fmsub_pd(__m128d __A, __m128d __B, __m128d __C)
 ///   32 bits.
 /// \returns A 128-bit vector of [4 x float] containing the result in the low
 ///    32 bits, and a copy of \a __A[127:32] in the upper 96 bits.
-static __inline__ __m128 __DEFAULT_FN_ATTRS128
-_mm_fmsub_ss(__m128 __A, __m128 __B, __m128 __C)
-{
-  return (__m128)__builtin_ia32_vfmaddss3((__v4sf)__A, (__v4sf)__B, -(__v4sf)__C);
+static __inline__ __m128 __DEFAULT_FN_ATTRS128_CONSTEXPR
+_mm_fmsub_ss(__m128 __A, __m128 __B, __m128 __C) {
+  __A[0] = __builtin_elementwise_fma(__A[0], __B[0], -__C[0]);
+  return __A;
 }
 
 /// Computes a scalar multiply-subtract of the double-precision values in
@@ -208,10 +224,10 @@ _mm_fmsub_ss(__m128 __A, __m128 __B, __m128 __C)
 ///    64 bits.
 /// \returns A 128-bit vector of [2 x double] containing the result in the low
 ///    64 bits, and a copy of \a __A[127:64] in the upper 64 bits.
-static __inline__ __m128d __DEFAULT_FN_ATTRS128
-_mm_fmsub_sd(__m128d __A, __m128d __B, __m128d __C)
-{
-  return (__m128d)__builtin_ia32_vfmaddsd3((__v2df)__A, (__v2df)__B, -(__v2df)__C);
+static __inline__ __m128d __DEFAULT_FN_ATTRS128_CONSTEXPR
+_mm_fmsub_sd(__m128d __A, __m128d __B, __m128d __C) {
+  __A[0] = __builtin_elementwise_fma(__A[0], __B[0], -__C[0]);
+  return __A;
 }
 
 /// Computes a negated multiply-add of 128-bit vectors of [4 x float].
@@ -228,10 +244,11 @@ _mm_fmsub_sd(__m128d __A, __m128d __B, __m128d __C)
 /// \param __C
 ///    A 128-bit vector of [4 x float] containing the addend.
 /// \returns A 128-bit [4 x float] vector containing the result.
-static __inline__ __m128 __DEFAULT_FN_ATTRS128
+static __inline__ __m128 __DEFAULT_FN_ATTRS128_CONSTEXPR
 _mm_fnmadd_ps(__m128 __A, __m128 __B, __m128 __C)
 {
-  return (__m128)__builtin_ia32_vfmaddps(-(__v4sf)__A, (__v4sf)__B, (__v4sf)__C);
+  return (__m128)__builtin_elementwise_fma(-(__v4sf)__A, (__v4sf)__B,
+                                           (__v4sf)__C);
 }
 
 /// Computes a negated multiply-add of 128-bit vectors of [2 x double].
@@ -248,10 +265,11 @@ _mm_fnmadd_ps(__m128 __A, __m128 __B, __m128 __C)
 /// \param __C
 ///    A 128-bit vector of [2 x double] containing the addend.
 /// \returns A 128-bit vector of [2 x double] containing the result.
-static __inline__ __m128d __DEFAULT_FN_ATTRS128
+static __inline__ __m128d __DEFAULT_FN_ATTRS128_CONSTEXPR
 _mm_fnmadd_pd(__m128d __A, __m128d __B, __m128d __C)
 {
-  return (__m128d)__builtin_ia32_vfmaddpd(-(__v2df)__A, (__v2df)__B, (__v2df)__C);
+  return (__m128d)__builtin_elementwise_fma(-(__v2df)__A, (__v2df)__B,
+                                            (__v2df)__C);
 }
 
 /// Computes a scalar negated multiply-add of the single-precision values in
@@ -277,10 +295,10 @@ _mm_fnmadd_pd(__m128d __A, __m128d __B, __m128d __C)
 ///    32 bits.
 /// \returns A 128-bit vector of [4 x float] containing the result in the low
 ///    32 bits, and a copy of \a __A[127:32] in the upper 96 bits.
-static __inline__ __m128 __DEFAULT_FN_ATTRS128
-_mm_fnmadd_ss(__m128 __A, __m128 __B, __m128 __C)
-{
-  return (__m128)__builtin_ia32_vfmaddss3((__v4sf)__A, -(__v4sf)__B, (__v4sf)__C);
+static __inline__ __m128 __DEFAULT_FN_ATTRS128_CONSTEXPR
+_mm_fnmadd_ss(__m128 __A, __m128 __B, __m128 __C) {
+  __A[0] = __builtin_elementwise_fma(__A[0], -__B[0], __C[0]);
+  return __A;
 }
 
 /// Computes a scalar negated multiply-add of the double-precision values
@@ -306,10 +324,10 @@ _mm_fnmadd_ss(__m128 __A, __m128 __B, __m128 __C)
 ///    64 bits.
 /// \returns A 128-bit vector of [2 x double] containing the result in the low
 ///    64 bits, and a copy of \a __A[127:64] in the upper 64 bits.
-static __inline__ __m128d __DEFAULT_FN_ATTRS128
-_mm_fnmadd_sd(__m128d __A, __m128d __B, __m128d __C)
-{
-  return (__m128d)__builtin_ia32_vfmaddsd3((__v2df)__A, -(__v2df)__B, (__v2df)__C);
+static __inline__ __m128d __DEFAULT_FN_ATTRS128_CONSTEXPR
+_mm_fnmadd_sd(__m128d __A, __m128d __B, __m128d __C) {
+  __A[0] = __builtin_elementwise_fma(__A[0], -__B[0], __C[0]);
+  return __A;
 }
 
 /// Computes a negated multiply-subtract of 128-bit vectors of [4 x float].
@@ -326,10 +344,11 @@ _mm_fnmadd_sd(__m128d __A, __m128d __B, __m128d __C)
 /// \param __C
 ///    A 128-bit vector of [4 x float] containing the subtrahend.
 /// \returns A 128-bit vector of [4 x float] containing the result.
-static __inline__ __m128 __DEFAULT_FN_ATTRS128
+static __inline__ __m128 __DEFAULT_FN_ATTRS128_CONSTEXPR
 _mm_fnmsub_ps(__m128 __A, __m128 __B, __m128 __C)
 {
-  return (__m128)__builtin_ia32_vfmaddps(-(__v4sf)__A, (__v4sf)__B, -(__v4sf)__C);
+  return (__m128)__builtin_elementwise_fma(-(__v4sf)__A, (__v4sf)__B,
+                                           -(__v4sf)__C);
 }
 
 /// Computes a negated multiply-subtract of 128-bit vectors of [2 x double].
@@ -346,10 +365,11 @@ _mm_fnmsub_ps(__m128 __A, __m128 __B, __m128 __C)
 /// \param __C
 ///    A 128-bit vector of [2 x double] containing the subtrahend.
 /// \returns A 128-bit vector of [2 x double] containing the result.
-static __inline__ __m128d __DEFAULT_FN_ATTRS128
+static __inline__ __m128d __DEFAULT_FN_ATTRS128_CONSTEXPR
 _mm_fnmsub_pd(__m128d __A, __m128d __B, __m128d __C)
 {
-  return (__m128d)__builtin_ia32_vfmaddpd(-(__v2df)__A, (__v2df)__B, -(__v2df)__C);
+  return (__m128d)__builtin_elementwise_fma(-(__v2df)__A, (__v2df)__B,
+                                            -(__v2df)__C);
 }
 
 /// Computes a scalar negated multiply-subtract of the single-precision
@@ -375,10 +395,10 @@ _mm_fnmsub_pd(__m128d __A, __m128d __B, __m128d __C)
 ///    32 bits.
 /// \returns A 128-bit vector of [4 x float] containing the result in the low
 ///    32 bits, and a copy of \a __A[127:32] in the upper 96 bits.
-static __inline__ __m128 __DEFAULT_FN_ATTRS128
-_mm_fnmsub_ss(__m128 __A, __m128 __B, __m128 __C)
-{
-  return (__m128)__builtin_ia32_vfmaddss3((__v4sf)__A, -(__v4sf)__B, -(__v4sf)__C);
+static __inline__ __m128 __DEFAULT_FN_ATTRS128_CONSTEXPR
+_mm_fnmsub_ss(__m128 __A, __m128 __B, __m128 __C) {
+  __A[0] = __builtin_elementwise_fma(__A[0], -__B[0], -__C[0]);
+  return __A;
 }
 
 /// Computes a scalar negated multiply-subtract of the double-precision
@@ -404,10 +424,10 @@ _mm_fnmsub_ss(__m128 __A, __m128 __B, __m128 __C)
 ///    64 bits.
 /// \returns A 128-bit vector of [2 x double] containing the result in the low
 ///    64 bits, and a copy of \a __A[127:64] in the upper 64 bits.
-static __inline__ __m128d __DEFAULT_FN_ATTRS128
-_mm_fnmsub_sd(__m128d __A, __m128d __B, __m128d __C)
-{
-  return (__m128d)__builtin_ia32_vfmaddsd3((__v2df)__A, -(__v2df)__B, -(__v2df)__C);
+static __inline__ __m128d __DEFAULT_FN_ATTRS128_CONSTEXPR
+_mm_fnmsub_sd(__m128d __A, __m128d __B, __m128d __C) {
+  __A[0] = __builtin_elementwise_fma(__A[0], -__B[0], -__C[0]);
+  return __A;
 }
 
 /// Computes a multiply with alternating add/subtract of 128-bit vectors of
@@ -528,10 +548,11 @@ _mm_fmsubadd_pd(__m128d __A, __m128d __B, __m128d __C)
 /// \param __C
 ///    A 256-bit vector of [8 x float] containing the addend.
 /// \returns A 256-bit vector of [8 x float] containing the result.
-static __inline__ __m256 __DEFAULT_FN_ATTRS256
+static __inline__ __m256 __DEFAULT_FN_ATTRS256_CONSTEXPR
 _mm256_fmadd_ps(__m256 __A, __m256 __B, __m256 __C)
 {
-  return (__m256)__builtin_ia32_vfmaddps256((__v8sf)__A, (__v8sf)__B, (__v8sf)__C);
+  return (__m256)__builtin_elementwise_fma((__v8sf)__A, (__v8sf)__B,
+                                           (__v8sf)__C);
 }
 
 /// Computes a multiply-add of 256-bit vectors of [4 x double].
@@ -548,10 +569,11 @@ _mm256_fmadd_ps(__m256 __A, __m256 __B, __m256 __C)
 /// \param __C
 ///    A 256-bit vector of [4 x double] containing the addend.
 /// \returns A 256-bit vector of [4 x double] containing the result.
-static __inline__ __m256d __DEFAULT_FN_ATTRS256
+static __inline__ __m256d __DEFAULT_FN_ATTRS256_CONSTEXPR
 _mm256_fmadd_pd(__m256d __A, __m256d __B, __m256d __C)
 {
-  return (__m256d)__builtin_ia32_vfmaddpd256((__v4df)__A, (__v4df)__B, (__v4df)__C);
+  return (__m256d)__builtin_elementwise_fma((__v4df)__A, (__v4df)__B,
+                                            (__v4df)__C);
 }
 
 /// Computes a multiply-subtract of 256-bit vectors of [8 x float].
@@ -568,10 +590,11 @@ _mm256_fmadd_pd(__m256d __A, __m256d __B, __m256d __C)
 /// \param __C
 ///    A 256-bit vector of [8 x float] containing the subtrahend.
 /// \returns A 256-bit vector of [8 x float] containing the result.
-static __inline__ __m256 __DEFAULT_FN_ATTRS256
+static __inline__ __m256 __DEFAULT_FN_ATTRS256_CONSTEXPR
 _mm256_fmsub_ps(__m256 __A, __m256 __B, __m256 __C)
 {
-  return (__m256)__builtin_ia32_vfmaddps256((__v8sf)__A, (__v8sf)__B, -(__v8sf)__C);
+  return (__m256)__builtin_elementwise_fma((__v8sf)__A, (__v8sf)__B,
+                                           -(__v8sf)__C);
 }
 
 /// Computes a multiply-subtract of 256-bit vectors of [4 x double].
@@ -588,10 +611,11 @@ _mm256_fmsub_ps(__m256 __A, __m256 __B, __m256 __C)
 /// \param __C
 ///    A 256-bit vector of [4 x double] containing the subtrahend.
 /// \returns A 256-bit vector of [4 x double] containing the result.
-static __inline__ __m256d __DEFAULT_FN_ATTRS256
+static __inline__ __m256d __DEFAULT_FN_ATTRS256_CONSTEXPR
 _mm256_fmsub_pd(__m256d __A, __m256d __B, __m256d __C)
 {
-  return (__m256d)__builtin_ia32_vfmaddpd256((__v4df)__A, (__v4df)__B, -(__v4df)__C);
+  return (__m256d)__builtin_elementwise_fma((__v4df)__A, (__v4df)__B,
+                                            -(__v4df)__C);
 }
 
 /// Computes a negated multiply-add of 256-bit vectors of [8 x float].
@@ -608,10 +632,11 @@ _mm256_fmsub_pd(__m256d __A, __m256d __B, __m256d __C)
 /// \param __C
 ///    A 256-bit vector of [8 x float] containing the addend.
 /// \returns A 256-bit vector of [8 x float] containing the result.
-static __inline__ __m256 __DEFAULT_FN_ATTRS256
+static __inline__ __m256 __DEFAULT_FN_ATTRS256_CONSTEXPR
 _mm256_fnmadd_ps(__m256 __A, __m256 __B, __m256 __C)
 {
-  return (__m256)__builtin_ia32_vfmaddps256(-(__v8sf)__A, (__v8sf)__B, (__v8sf)__C);
+  return (__m256)__builtin_elementwise_fma(-(__v8sf)__A, (__v8sf)__B,
+                                           (__v8sf)__C);
 }
 
 /// Computes a negated multiply-add of 256-bit vectors of [4 x double].
@@ -628,10 +653,11 @@ _mm256_fnmadd_ps(__m256 __A, __m256 __B, __m256 __C)
 /// \param __C
 ///    A 256-bit vector of [4 x double] containing the addend.
 /// \returns A 256-bit vector of [4 x double] containing the result.
-static __inline__ __m256d __DEFAULT_FN_ATTRS256
+static __inline__ __m256d __DEFAULT_FN_ATTRS256_CONSTEXPR
 _mm256_fnmadd_pd(__m256d __A, __m256d __B, __m256d __C)
 {
-  return (__m256d)__builtin_ia32_vfmaddpd256(-(__v4df)__A, (__v4df)__B, (__v4df)__C);
+  return (__m256d)__builtin_elementwise_fma(-(__v4df)__A, (__v4df)__B,
+                                            (__v4df)__C);
 }
 
 /// Computes a negated multiply-subtract of 256-bit vectors of [8 x float].
@@ -648,10 +674,11 @@ _mm256_fnmadd_pd(__m256d __A, __m256d __B, __m256d __C)
 /// \param __C
 ///    A 256-bit vector of [8 x float] containing the subtrahend.
 /// \returns A 256-bit vector of [8 x float] containing the result.
-static __inline__ __m256 __DEFAULT_FN_ATTRS256
+static __inline__ __m256 __DEFAULT_FN_ATTRS256_CONSTEXPR
 _mm256_fnmsub_ps(__m256 __A, __m256 __B, __m256 __C)
 {
-  return (__m256)__builtin_ia32_vfmaddps256(-(__v8sf)__A, (__v8sf)__B, -(__v8sf)__C);
+  return (__m256)__builtin_elementwise_fma(-(__v8sf)__A, (__v8sf)__B,
+                                           -(__v8sf)__C);
 }
 
 /// Computes a negated multiply-subtract of 256-bit vectors of [4 x double].
@@ -668,10 +695,11 @@ _mm256_fnmsub_ps(__m256 __A, __m256 __B, __m256 __C)
 /// \param __C
 ///    A 256-bit vector of [4 x double] containing the subtrahend.
 /// \returns A 256-bit vector of [4 x double] containing the result.
-static __inline__ __m256d __DEFAULT_FN_ATTRS256
+static __inline__ __m256d __DEFAULT_FN_ATTRS256_CONSTEXPR
 _mm256_fnmsub_pd(__m256d __A, __m256d __B, __m256d __C)
 {
-  return (__m256d)__builtin_ia32_vfmaddpd256(-(__v4df)__A, (__v4df)__B, -(__v4df)__C);
+  return (__m256d)__builtin_elementwise_fma(-(__v4df)__A, (__v4df)__B,
+                                            -(__v4df)__C);
 }
 
 /// Computes a multiply with alternating add/subtract of 256-bit vectors of
@@ -792,5 +820,7 @@ _mm256_fmsubadd_pd(__m256d __A, __m256d __B, __m256d __C)
 
 #undef __DEFAULT_FN_ATTRS128
 #undef __DEFAULT_FN_ATTRS256
+#undef __DEFAULT_FN_ATTRS128_CONSTEXPR
+#undef __DEFAULT_FN_ATTRS256_CONSTEXPR
 
 #endif /* __FMAINTRIN_H */

package/lib/clang/{20 → 22}/include/gfniintrin.h

@@ -14,29 +14,36 @@
 #ifndef __GFNIINTRIN_H
 #define __GFNIINTRIN_H
 
-#if defined(__EVEX512__) && !defined(__AVX10_1_512__)
 /* Default attributes for simple form (no masking). */
+#if defined(__cplusplus) && (__cplusplus >= 201103L)
 #define __DEFAULT_FN_ATTRS                                                     \
-  __attribute__((__always_inline__, __nodebug__,                               \
-                 __target__("gfni,no-evex512"), __min_vector_width__(128)))
+  __attribute__((__always_inline__, __nodebug__, __target__("gfni"),           \
+                 __min_vector_width__(128))) constexpr
 
 /* Default attributes for YMM unmasked form. */
 #define __DEFAULT_FN_ATTRS_Y                                                   \
-  __attribute__((__always_inline__, __nodebug__,                               \
-                 __target__("avx,gfni,no-evex512"),                            \
-                 __min_vector_width__(256)))
+  __attribute__((__always_inline__, __nodebug__, __target__("avx,gfni"),       \
+                 __min_vector_width__(256))) constexpr
 
 /* Default attributes for VLX masked forms. */
 #define __DEFAULT_FN_ATTRS_VL128                                               \
   __attribute__((__always_inline__, __nodebug__,                               \
-                 __target__("avx512bw,avx512vl,gfni,no-evex512"),              \
-                 __min_vector_width__(128)))
+                 __target__("avx512bw,avx512vl,gfni"),                         \
+                 __min_vector_width__(128))) constexpr
 #define __DEFAULT_FN_ATTRS_VL256                                               \
   __attribute__((__always_inline__, __nodebug__,                               \
-                 __target__("avx512bw,avx512vl,gfni,no-evex512"),              \
-                 __min_vector_width__(256)))
+                 __target__("avx512bw,avx512vl,gfni"),                         \
+                 __min_vector_width__(256))) constexpr
+
+/* Default attributes for ZMM unmasked forms. */
+#define __DEFAULT_FN_ATTRS_Z                                                   \
+  __attribute__((__always_inline__, __nodebug__, __target__("avx512f,gfni"),   \
+                 __min_vector_width__(512))) constexpr
+/* Default attributes for ZMM masked forms. */
+#define __DEFAULT_FN_ATTRS_Z_MASK                                              \
+  __attribute__((__always_inline__, __nodebug__, __target__("avx512bw,gfni"),  \
+                 __min_vector_width__(512))) constexpr
 #else
-/* Default attributes for simple form (no masking). */
 #define __DEFAULT_FN_ATTRS                                                     \
   __attribute__((__always_inline__, __nodebug__, __target__("gfni"),           \
                  __min_vector_width__(128)))
@@ -55,18 +62,16 @@
   __attribute__((__always_inline__, __nodebug__,                               \
                  __target__("avx512bw,avx512vl,gfni"),                         \
                  __min_vector_width__(256)))
-#endif
 
 /* Default attributes for ZMM unmasked forms. */
 #define __DEFAULT_FN_ATTRS_Z                                                   \
-  __attribute__((__always_inline__, __nodebug__,                               \
-                 __target__("avx512f,evex512,gfni"),                           \
+  __attribute__((__always_inline__, __nodebug__, __target__("avx512f,gfni"),   \
                  __min_vector_width__(512)))
 /* Default attributes for ZMM masked forms. */
 #define __DEFAULT_FN_ATTRS_Z_MASK                                              \
-  __attribute__((__always_inline__, __nodebug__,                               \
-                 __target__("avx512bw,evex512,gfni"),                          \
+  __attribute__((__always_inline__, __nodebug__, __target__("avx512bw,gfni"),  \
                  __min_vector_width__(512)))
+#endif
 
 #define _mm_gf2p8affineinv_epi64_epi8(A, B, I) \
   ((__m128i)__builtin_ia32_vgf2p8affineinvqb_v16qi((__v16qi)(__m128i)(A), \

package/lib/clang/{20 → 22}/include/gpuintrin.h

@@ -25,21 +25,47 @@
 #endif
 #endif
 
+#include <stdint.h>
+
+#if !defined(__cplusplus)
+_Pragma("push_macro(\"bool\")");
+#define bool _Bool
+#endif
+
+_Pragma("omp begin declare target device_type(nohost)");
+_Pragma("omp begin declare variant match(device = {kind(gpu)})");
+
+// Forward declare a few functions for the implementation header.
+
+// Returns a bitmask marking all lanes that have the same value of __x.
+_DEFAULT_FN_ATTRS static __inline__ uint64_t
+__gpu_match_any_u32_impl(uint64_t __lane_mask, uint32_t __x);
+
+// Returns a bitmask marking all lanes that have the same value of __x.
+_DEFAULT_FN_ATTRS static __inline__ uint64_t
+__gpu_match_any_u64_impl(uint64_t __lane_mask, uint64_t __x);
+
+// Returns the current lane mask if every lane contains __x.
+_DEFAULT_FN_ATTRS static __inline__ uint64_t
+__gpu_match_all_u32_impl(uint64_t __lane_mask, uint32_t __x);
+
+// Returns the current lane mask if every lane contains __x.
+_DEFAULT_FN_ATTRS static __inline__ uint64_t
+__gpu_match_all_u64_impl(uint64_t __lane_mask, uint64_t __x);
+
+_Pragma("omp end declare variant");
+_Pragma("omp end declare target");
+
 #if defined(__NVPTX__)
 #include <nvptxintrin.h>
 #elif defined(__AMDGPU__)
 #include <amdgpuintrin.h>
+#elif defined(__SPIRV__)
+#include <spirvintrin.h>
 #elif !defined(_OPENMP)
 #error "This header is only meant to be used on GPU architectures."
 #endif
 
-#include <stdint.h>
-
-#if !defined(__cplusplus)
-_Pragma("push_macro(\"bool\")");
-#define bool _Bool
-#endif
-
 _Pragma("omp begin declare target device_type(nohost)");
 _Pragma("omp begin declare variant match(device = {kind(gpu)})");
 
@@ -57,7 +83,7 @@ _DEFAULT_FN_ATTRS static __inline__ uint32_t __gpu_num_blocks(int __dim) {
   case 2:
     return __gpu_num_blocks_z();
   default:
-    __builtin_unreachable();
+    return 1;
   }
 }
 
@@ -71,7 +97,7 @@ _DEFAULT_FN_ATTRS static __inline__ uint32_t __gpu_block_id(int __dim) {
   case 2:
     return __gpu_block_id_z();
   default:
-    __builtin_unreachable();
+    return 0;
   }
 }
 
@@ -85,7 +111,7 @@ _DEFAULT_FN_ATTRS static __inline__ uint32_t __gpu_num_threads(int __dim) {
   case 2:
     return __gpu_num_threads_z();
   default:
-    __builtin_unreachable();
+    return 1;
   }
 }
 
@@ -99,7 +125,7 @@ _DEFAULT_FN_ATTRS static __inline__ uint32_t __gpu_thread_id(int __dim) {
   case 2:
     return __gpu_thread_id_z();
   default:
-    __builtin_unreachable();
+    return 0;
   }
 }
 
@@ -115,6 +141,16 @@ __gpu_is_first_in_lane(uint64_t __lane_mask) {
   return __gpu_lane_id() == __gpu_first_lane_id(__lane_mask);
 }
 
+// Copies the value from the first active thread to the rest.
+_DEFAULT_FN_ATTRS static __inline__ uint64_t
+__gpu_read_first_lane_u64(uint64_t __lane_mask, uint64_t __x) {
+  uint32_t __hi = (uint32_t)(__x >> 32ull);
+  uint32_t __lo = (uint32_t)(__x & 0xFFFFFFFFull);
+  return ((uint64_t)__gpu_read_first_lane_u32(__lane_mask, __hi) << 32ull) |
+         ((uint64_t)__gpu_read_first_lane_u32(__lane_mask, __lo) &
+          0xFFFFFFFFull);
+}
+
 // Gets the first floating point value from the active lanes.
 _DEFAULT_FN_ATTRS static __inline__ float
 __gpu_read_first_lane_f32(uint64_t __lane_mask, float __x) {
@@ -131,6 +167,18 @@ __gpu_read_first_lane_f64(uint64_t __lane_mask, double __x) {
                                         __builtin_bit_cast(uint64_t, __x)));
 }
 
+// Shuffles the the lanes according to the given index.
+_DEFAULT_FN_ATTRS static __inline__ uint64_t
+__gpu_shuffle_idx_u64(uint64_t __lane_mask, uint32_t __idx, uint64_t __x,
+                      uint32_t __width) {
+  uint32_t __hi = (uint32_t)(__x >> 32ull);
+  uint32_t __lo = (uint32_t)(__x & 0xFFFFFFFF);
+  uint32_t __mask = (uint32_t)__lane_mask;
+  return ((uint64_t)__gpu_shuffle_idx_u32(__mask, __idx, __hi, __width)
+          << 32ull) |
+         ((uint64_t)__gpu_shuffle_idx_u32(__mask, __idx, __lo, __width));
+}
+
 // Shuffles the the lanes according to the given index.
 _DEFAULT_FN_ATTRS static __inline__ float
 __gpu_shuffle_idx_f32(uint64_t __lane_mask, uint32_t __idx, float __x,
@@ -212,6 +260,64 @@ __DO_LANE_SUM(float, f32);    // float __gpu_lane_sum_f32(m, x)
 __DO_LANE_SUM(double, f64);   // double __gpu_lane_sum_f64(m, x)
 #undef __DO_LANE_SUM
 
+// Returns a bitmask marking all lanes that have the same value of __x.
+_DEFAULT_FN_ATTRS static __inline__ uint64_t
+__gpu_match_any_u32_impl(uint64_t __lane_mask, uint32_t __x) {
+  uint64_t __match_mask = 0;
+
+  bool __done = 0;
+  for (uint64_t __active_mask = __lane_mask; __active_mask;
+       __active_mask = __gpu_ballot(__lane_mask, !__done)) {
+    if (!__done) {
+      uint32_t __first = __gpu_read_first_lane_u32(__active_mask, __x);
+      if (__first == __x) {
+        __match_mask = __gpu_lane_mask();
+        __done = 1;
+      }
+    }
+  }
+  __gpu_sync_lane(__lane_mask);
+  return __match_mask;
+}
+
+// Returns a bitmask marking all lanes that have the same value of __x.
+_DEFAULT_FN_ATTRS static __inline__ uint64_t
+__gpu_match_any_u64_impl(uint64_t __lane_mask, uint64_t __x) {
+  uint64_t __match_mask = 0;
+
+  bool __done = 0;
+  for (uint64_t __active_mask = __lane_mask; __active_mask;
+       __active_mask = __gpu_ballot(__lane_mask, !__done)) {
+    if (!__done) {
+      uint64_t __first = __gpu_read_first_lane_u64(__active_mask, __x);
+      if (__first == __x) {
+        __match_mask = __gpu_lane_mask();
+        __done = 1;
+      }
+    }
+  }
+  __gpu_sync_lane(__lane_mask);
+  return __match_mask;
+}
+
+// Returns the current lane mask if every lane contains __x.
+_DEFAULT_FN_ATTRS static __inline__ uint64_t
+__gpu_match_all_u32_impl(uint64_t __lane_mask, uint32_t __x) {
+  uint32_t __first = __gpu_read_first_lane_u32(__lane_mask, __x);
+  uint64_t __ballot = __gpu_ballot(__lane_mask, __x == __first);
+  __gpu_sync_lane(__lane_mask);
+  return __ballot == __gpu_lane_mask() ? __gpu_lane_mask() : 0ull;
+}
+
+// Returns the current lane mask if every lane contains __x.
+_DEFAULT_FN_ATTRS static __inline__ uint64_t
+__gpu_match_all_u64_impl(uint64_t __lane_mask, uint64_t __x) {
+  uint64_t __first = __gpu_read_first_lane_u64(__lane_mask, __x);
+  uint64_t __ballot = __gpu_ballot(__lane_mask, __x == __first);
+  __gpu_sync_lane(__lane_mask);
+  return __ballot == __gpu_lane_mask() ? __gpu_lane_mask() : 0ull;
+}
+
 _Pragma("omp end declare variant");
 _Pragma("omp end declare target");