404 Not Found

PATH: lib64 / llvm17 / lib / clang / 17 / include

/*===---- bmi2intrin.h - BMI2 intrinsics -----------------------------------===
 *
 * Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 * See https://llvm.org/LICENSE.txt for license information.
 * SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 *
 *===-----------------------------------------------------------------------===
 */

#ifndef __IMMINTRIN_H
#error "Never use <bmi2intrin.h> directly; include <immintrin.h> instead."
#endif

#ifndef __BMI2INTRIN_H
#define __BMI2INTRIN_H

/* Define the default attributes for the functions in this file. */
#define __DEFAULT_FN_ATTRS __attribute__((__always_inline__, __nodebug__, __target__("bmi2")))

/// Copies the unsigned 32-bit integer \a __X and zeroes the upper bits
///    starting at bit number \a __Y.
///
/// \code{.operation}
/// i := __Y[7:0]
/// result := __X
/// IF i < 32
///   result[31:i] := 0
/// FI
/// \endcode
///
/// \headerfile <immintrin.h>
///
/// This intrinsic corresponds to the \c BZHI instruction.
///
/// \param __X
///    The 32-bit source value to copy.
/// \param __Y
///    The lower 8 bits specify the bit number of the lowest bit to zero.
/// \returns The partially zeroed 32-bit value.
static __inline__ unsigned int __DEFAULT_FN_ATTRS
_bzhi_u32(unsigned int __X, unsigned int __Y)
{
  return __builtin_ia32_bzhi_si(__X, __Y);
}

/// Deposit (scatter) low-order bits from the unsigned 32-bit integer \a __X
///    into the 32-bit result, according to the mask in the unsigned 32-bit
///    integer \a __Y. All other bits of the result are zero.
///
/// \code{.operation}
/// i := 0
/// result := 0
/// FOR m := 0 TO 31
///   IF __Y[m] == 1
///     result[m] := __X[i]
///     i := i + 1
///   ENDIF
/// ENDFOR
/// \endcode
///
/// \headerfile <immintrin.h>
///
/// This intrinsic corresponds to the \c PDEP instruction.
///
/// \param __X
///    The 32-bit source value to copy.
/// \param __Y
///    The 32-bit mask specifying where to deposit source bits.
/// \returns The 32-bit result.
static __inline__ unsigned int __DEFAULT_FN_ATTRS
_pdep_u32(unsigned int __X, unsigned int __Y)
{
  return __builtin_ia32_pdep_si(__X, __Y);
}

/// Extract (gather) bits from the unsigned 32-bit integer \a __X into the
///    low-order bits of the 32-bit result, according to the mask in the
///    unsigned 32-bit integer \a __Y. All other bits of the result are zero.
///
/// \code{.operation}
/// i := 0
/// result := 0
/// FOR m := 0 TO 31
///   IF __Y[m] == 1
///     result[i] := __X[m]
///     i := i + 1
///   ENDIF
/// ENDFOR
/// \endcode
///
/// \headerfile <immintrin.h>
///
/// This intrinsic corresponds to the \c PEXT instruction.
///
/// \param __X
///    The 32-bit source value to copy.
/// \param __Y
///    The 32-bit mask specifying which source bits to extract.
/// \returns The 32-bit result.
static __inline__ unsigned int __DEFAULT_FN_ATTRS
_pext_u32(unsigned int __X, unsigned int __Y)
{
  return __builtin_ia32_pext_si(__X, __Y);
}

/// Multiplies the unsigned 32-bit integers \a __X and \a __Y to form a
///    64-bit product. Stores the upper 32 bits of the product in the
///    memory at \a __P and returns the lower 32 bits.
///
/// \code{.operation}
/// Store32(__P, (__X * __Y)[63:32])
/// result := (__X * __Y)[31:0]
/// \endcode
///
/// \headerfile <immintrin.h>
///
/// This intrinsic corresponds to the \c MULX instruction.
///
/// \param __X
///    An unsigned 32-bit multiplicand.
/// \param __Y
///    An unsigned 32-bit multiplicand.
/// \param __P
///    A pointer to memory for storing the upper half of the product.
/// \returns The lower half of the product.
static __inline__ unsigned int __DEFAULT_FN_ATTRS
_mulx_u32(unsigned int __X, unsigned int __Y, unsigned int *__P)
{
  unsigned long long __res = (unsigned long long) __X * __Y;
  *__P = (unsigned int)(__res >> 32);
  return (unsigned int)__res;
}

#ifdef  __x86_64__

/// Copies the unsigned 64-bit integer \a __X and zeroes the upper bits
///    starting at bit number \a __Y.
///
/// \code{.operation}
/// i := __Y[7:0]
/// result := __X
/// IF i < 64
///   result[63:i] := 0
/// FI
/// \endcode
///
/// \headerfile <immintrin.h>
///
/// This intrinsic corresponds to the \c BZHI instruction.
///
/// \param __X
///    The 64-bit source value to copy.
/// \param __Y
///    The lower 8 bits specify the bit number of the lowest bit to zero.
/// \returns The partially zeroed 64-bit value.
static __inline__ unsigned long long __DEFAULT_FN_ATTRS
_bzhi_u64(unsigned long long __X, unsigned long long __Y)
{
  return __builtin_ia32_bzhi_di(__X, __Y);
}

/// Deposit (scatter) low-order bits from the unsigned 64-bit integer \a __X
///    into the 64-bit result, according to the mask in the unsigned 64-bit
///    integer \a __Y. All other bits of the result are zero.
///
/// \code{.operation}
/// i := 0
/// result := 0
/// FOR m := 0 TO 63
///   IF __Y[m] == 1
///     result[m] := __X[i]
///     i := i + 1
///   ENDIF
/// ENDFOR
/// \endcode
///
/// \headerfile <immintrin.h>
///
/// This intrinsic corresponds to the \c PDEP instruction.
///
/// \param __X
///    The 64-bit source value to copy.
/// \param __Y
///    The 64-bit mask specifying where to deposit source bits.
/// \returns The 64-bit result.
static __inline__ unsigned long long __DEFAULT_FN_ATTRS
_pdep_u64(unsigned long long __X, unsigned long long __Y)
{
  return __builtin_ia32_pdep_di(__X, __Y);
}

/// Extract (gather) bits from the unsigned 64-bit integer \a __X into the
///    low-order bits of the 64-bit result, according to the mask in the
///    unsigned 64-bit integer \a __Y. All other bits of the result are zero.
///
/// \code{.operation}
/// i := 0
/// result := 0
/// FOR m := 0 TO 63
///   IF __Y[m] == 1
///     result[i] := __X[m]
///     i := i + 1
///   ENDIF
/// ENDFOR
/// \endcode
///
/// \headerfile <immintrin.h>
///
/// This intrinsic corresponds to the \c PEXT instruction.
///
/// \param __X
///    The 64-bit source value to copy.
/// \param __Y
///    The 64-bit mask specifying which source bits to extract.
/// \returns The 64-bit result.
static __inline__ unsigned long long __DEFAULT_FN_ATTRS
_pext_u64(unsigned long long __X, unsigned long long __Y)
{
  return __builtin_ia32_pext_di(__X, __Y);
}

/// Multiplies the unsigned 64-bit integers \a __X and \a __Y to form a
///    128-bit product. Stores the upper 64 bits of the product to the
///    memory addressed by \a __P and returns the lower 64 bits.
///
/// \code{.operation}
/// Store64(__P, (__X * __Y)[127:64])
/// result := (__X * __Y)[63:0]
/// \endcode
///
/// \headerfile <immintrin.h>
///
/// This intrinsic corresponds to the \c MULX instruction.
///
/// \param __X
///    An unsigned 64-bit multiplicand.
/// \param __Y
///    An unsigned 64-bit multiplicand.
/// \param __P
///    A pointer to memory for storing the upper half of the product.
/// \returns The lower half of the product.
static __inline__ unsigned long long __DEFAULT_FN_ATTRS
_mulx_u64 (unsigned long long __X, unsigned long long __Y,
	   unsigned long long *__P)
{
  unsigned __int128 __res = (unsigned __int128) __X * __Y;
  *__P = (unsigned long long) (__res >> 64);
  return (unsigned long long) __res;
}

#endif /* __x86_64__  */

#undef __DEFAULT_FN_ATTRS

#endif /* __BMI2INTRIN_H */

[+] ..
[-] riscv_ntlh.h [edit]
[-] avx512vlbitalgintrin.h [edit]
[-] avxifmaintrin.h [edit]
[-] mm_malloc.h [edit]
[-] avx512vlbf16intrin.h [edit]
[-] mwaitxintrin.h [edit]
[-] adxintrin.h [edit]
[-] arm_cmse.h [edit]
[-] __clang_cuda_intrinsics.h [edit]
[-] module.modulemap [edit]
[-] crc32intrin.h [edit]
[+] cuda_wrappers
[-] msa.h [edit]
[-] __clang_hip_stdlib.h [edit]
[-] arm_acle.h [edit]
[-] arm_neon_sve_bridge.h [edit]
[-] xsaveintrin.h [edit]
[-] velintrin_gen.h [edit]
[-] stdatomic.h [edit]
[-] hvx_hexagon_protos.h [edit]
[-] uintrintrin.h [edit]
[-] fxsrintrin.h [edit]
[-] sha512intrin.h [edit]
[-] __clang_cuda_texture_intrinsics.h [edit]
[-] iso646.h [edit]
[-] unwind.h [edit]
[-] avx512vlbwintrin.h [edit]
[-] avxvnniint8intrin.h [edit]
[-] avx512erintrin.h [edit]
[-] avxvnniintrin.h [edit]
[-] ia32intrin.h [edit]
[-] rdseedintrin.h [edit]
[-] prfchiintrin.h [edit]
[-] amxcomplexintrin.h [edit]
[-] clflushoptintrin.h [edit]
[-] htmxlintrin.h [edit]
[-] avx512fintrin.h [edit]
[-] gfniintrin.h [edit]
[-] arm_cde.h [edit]
[+] ppc_wrappers
[-] amxfp16intrin.h [edit]
[-] velintrin_approx.h [edit]
[-] avx512pfintrin.h [edit]
[-] stdarg.h [edit]
[-] cmpccxaddintrin.h [edit]
[-] avx512vlintrin.h [edit]
[-] __clang_hip_math.h [edit]
[-] vecintrin.h [edit]
[-] xtestintrin.h [edit]
[-] __wmmintrin_aes.h [edit]
[-] arm_neon.h [edit]
[-] immintrin.h [edit]
[-] ammintrin.h [edit]
[-] waitpkgintrin.h [edit]
[-] vpclmulqdqintrin.h [edit]
[-] fmaintrin.h [edit]
[-] tsxldtrkintrin.h [edit]
[-] prfchwintrin.h [edit]
[-] avx512bitalgintrin.h [edit]
[-] bmiintrin.h [edit]
[-] __wmmintrin_pclmul.h [edit]
[-] htmintrin.h [edit]
[-] mm3dnow.h [edit]
[-] __clang_cuda_builtin_vars.h [edit]
[-] __clang_hip_runtime_wrapper.h [edit]
[-] stdbool.h [edit]
[-] altivec.h [edit]
[-] wbnoinvdintrin.h [edit]
[-] keylockerintrin.h [edit]
[-] tgmath.h [edit]
[-] hexagon_circ_brev_intrinsics.h [edit]
[-] x86intrin.h [edit]
[-] pkuintrin.h [edit]
[-] avx512vbmivlintrin.h [edit]
[-] avxneconvertintrin.h [edit]
[-] __clang_hip_cmath.h [edit]
[-] sgxintrin.h [edit]
[-] f16cintrin.h [edit]
[-] opencl-c-base.h [edit]
[-] cpuid.h [edit]
[-] raointintrin.h [edit]
[-] builtins.h [edit]
[-] emmintrin.h [edit]
[-] smmintrin.h [edit]
[-] vaesintrin.h [edit]
[-] larchintrin.h [edit]
[-] avx512ifmaintrin.h [edit]
[-] intrin.h [edit]
[-] avx512vlvp2intersectintrin.h [edit]
[-] fma4intrin.h [edit]
[-] pmmintrin.h [edit]
[-] __clang_hip_libdevice_declares.h [edit]
[-] limits.h [edit]
[-] clwbintrin.h [edit]
[-] rtmintrin.h [edit]
[-] mmintrin.h [edit]
[-] stddef.h [edit]
[-] invpcidintrin.h [edit]
[-] avx512vp2intersectintrin.h [edit]
[-] cet.h [edit]
[-] xopintrin.h [edit]
[-] avx512vlvnniintrin.h [edit]
[-] avx512vlfp16intrin.h [edit]
[-] stdint.h [edit]
[-] arm64intr.h [edit]
[-] sm4intrin.h [edit]
[-] avx512vnniintrin.h [edit]
[-] avx2intrin.h [edit]
[-] movdirintrin.h [edit]
[-] tbmintrin.h [edit]
[-] arm_mve.h [edit]
[-] avx512ifmavlintrin.h [edit]
[-] amxintrin.h [edit]
[-] opencl-c.h [edit]
[-] stdalign.h [edit]
[-] __clang_cuda_device_functions.h [edit]
[-] pconfigintrin.h [edit]
[-] avx512fp16intrin.h [edit]
[-] inttypes.h [edit]
[-] arm_bf16.h [edit]
[-] __clang_cuda_math_forward_declares.h [edit]
[-] vadefs.h [edit]
[-] shaintrin.h [edit]
[-] hexagon_protos.h [edit]
[-] ptwriteintrin.h [edit]
[-] xsaveoptintrin.h [edit]
[-] enqcmdintrin.h [edit]
[-] x86gprintrin.h [edit]
[-] tmmintrin.h [edit]
[-] stdnoreturn.h [edit]
[-] avx512bf16intrin.h [edit]
[-] varargs.h [edit]
[-] s390intrin.h [edit]
[-] avx512vbmiintrin.h [edit]
[+] openmp_wrappers
[-] wmmintrin.h [edit]
[-] __clang_cuda_cmath.h [edit]
[-] clzerointrin.h [edit]
[-] xsavesintrin.h [edit]
[-] __clang_cuda_libdevice_declares.h [edit]
[-] nmmintrin.h [edit]
[-] wasm_simd128.h [edit]
[-] xsavecintrin.h [edit]
[-] avx512dqintrin.h [edit]
[-] lwpintrin.h [edit]
[-] serializeintrin.h [edit]
[-] arm_sme_draft_spec_subject_to_change.h [edit]
[-] avxintrin.h [edit]
[-] __stddef_max_align_t.h [edit]
[-] sm3intrin.h [edit]
[-] velintrin.h [edit]
[-] __clang_cuda_complex_builtins.h [edit]
[-] armintr.h [edit]
[-] avx512cdintrin.h [edit]
[-] float.h [edit]
[-] avx512vbmi2intrin.h [edit]
[-] lzcntintrin.h [edit]
[-] sifive_vector.h [edit]
[-] rdpruintrin.h [edit]
[-] arm_sve.h [edit]
[-] avx512vpopcntdqintrin.h [edit]
[-] xmmintrin.h [edit]
[-] hresetintrin.h [edit]
[-] bmi2intrin.h [edit]
[-] hexagon_types.h [edit]
[-] avx512bwintrin.h [edit]
[-] cetintrin.h [edit]
[-] __clang_cuda_math.h [edit]
[-] avx512vlvbmi2intrin.h [edit]
[-] arm_fp16.h [edit]
[-] avx512vpopcntdqvlintrin.h [edit]
[-] avxvnniint16intrin.h [edit]
[+] llvm_libc_wrappers
[-] __clang_cuda_runtime_wrapper.h [edit]
[-] cldemoteintrin.h [edit]
[-] avx512vldqintrin.h [edit]
[-] avx512vlcdintrin.h [edit]
[-] popcntintrin.h [edit]