clang  14.0.0git
Macros | Functions
tmmintrin.h File Reference
#include <pmmintrin.h>
Include dependency graph for tmmintrin.h:
This graph shows which files directly or indirectly include this file:

Go to the source code of this file.

Macros

#define __DEFAULT_FN_ATTRS   __attribute__((__always_inline__, __nodebug__, __target__("ssse3"), __min_vector_width__(64)))
 
#define __DEFAULT_FN_ATTRS_MMX   __attribute__((__always_inline__, __nodebug__, __target__("mmx,ssse3"), __min_vector_width__(64)))
 
#define _mm_alignr_epi8(a, b, n)
 Concatenates the two 128-bit integer vector operands, and right-shifts the result by the number of bytes specified in the immediate operand. More...
 
#define _mm_alignr_pi8(a, b, n)   ((__m64)__builtin_ia32_palignr((__v8qi)(__m64)(a), (__v8qi)(__m64)(b), (n)))
 Concatenates the two 64-bit integer vector operands, and right-shifts the result by the number of bytes specified in the immediate operand. More...
 

Functions

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_abs_pi8 (__m64 __a)
 Computes the absolute value of each of the packed 8-bit signed integers in the source operand and stores the 8-bit unsigned integer results in the destination. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_abs_epi8 (__m128i __a)
 Computes the absolute value of each of the packed 8-bit signed integers in the source operand and stores the 8-bit unsigned integer results in the destination. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_abs_pi16 (__m64 __a)
 Computes the absolute value of each of the packed 16-bit signed integers in the source operand and stores the 16-bit unsigned integer results in the destination. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_abs_epi16 (__m128i __a)
 Computes the absolute value of each of the packed 16-bit signed integers in the source operand and stores the 16-bit unsigned integer results in the destination. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_abs_pi32 (__m64 __a)
 Computes the absolute value of each of the packed 32-bit signed integers in the source operand and stores the 32-bit unsigned integer results in the destination. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_abs_epi32 (__m128i __a)
 Computes the absolute value of each of the packed 32-bit signed integers in the source operand and stores the 32-bit unsigned integer results in the destination. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_hadd_epi16 (__m128i __a, __m128i __b)
 Horizontally adds the adjacent pairs of values contained in 2 packed 128-bit vectors of [8 x i16]. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_hadd_epi32 (__m128i __a, __m128i __b)
 Horizontally adds the adjacent pairs of values contained in 2 packed 128-bit vectors of [4 x i32]. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_hadd_pi16 (__m64 __a, __m64 __b)
 Horizontally adds the adjacent pairs of values contained in 2 packed 64-bit vectors of [4 x i16]. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_hadd_pi32 (__m64 __a, __m64 __b)
 Horizontally adds the adjacent pairs of values contained in 2 packed 64-bit vectors of [2 x i32]. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_hadds_epi16 (__m128i __a, __m128i __b)
 Horizontally adds the adjacent pairs of values contained in 2 packed 128-bit vectors of [8 x i16]. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_hadds_pi16 (__m64 __a, __m64 __b)
 Horizontally adds the adjacent pairs of values contained in 2 packed 64-bit vectors of [4 x i16]. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_hsub_epi16 (__m128i __a, __m128i __b)
 Horizontally subtracts the adjacent pairs of values contained in 2 packed 128-bit vectors of [8 x i16]. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_hsub_epi32 (__m128i __a, __m128i __b)
 Horizontally subtracts the adjacent pairs of values contained in 2 packed 128-bit vectors of [4 x i32]. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_hsub_pi16 (__m64 __a, __m64 __b)
 Horizontally subtracts the adjacent pairs of values contained in 2 packed 64-bit vectors of [4 x i16]. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_hsub_pi32 (__m64 __a, __m64 __b)
 Horizontally subtracts the adjacent pairs of values contained in 2 packed 64-bit vectors of [2 x i32]. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_hsubs_epi16 (__m128i __a, __m128i __b)
 Horizontally subtracts the adjacent pairs of values contained in 2 packed 128-bit vectors of [8 x i16]. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_hsubs_pi16 (__m64 __a, __m64 __b)
 Horizontally subtracts the adjacent pairs of values contained in 2 packed 64-bit vectors of [4 x i16]. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_maddubs_epi16 (__m128i __a, __m128i __b)
 Multiplies corresponding pairs of packed 8-bit unsigned integer values contained in the first source operand and packed 8-bit signed integer values contained in the second source operand, adds pairs of contiguous products with signed saturation, and writes the 16-bit sums to the corresponding bits in the destination. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_maddubs_pi16 (__m64 __a, __m64 __b)
 Multiplies corresponding pairs of packed 8-bit unsigned integer values contained in the first source operand and packed 8-bit signed integer values contained in the second source operand, adds pairs of contiguous products with signed saturation, and writes the 16-bit sums to the corresponding bits in the destination. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_mulhrs_epi16 (__m128i __a, __m128i __b)
 Multiplies packed 16-bit signed integer values, truncates the 32-bit products to the 18 most significant bits by right-shifting, rounds the truncated value by adding 1, and writes bits [16:1] to the destination. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_mulhrs_pi16 (__m64 __a, __m64 __b)
 Multiplies packed 16-bit signed integer values, truncates the 32-bit products to the 18 most significant bits by right-shifting, rounds the truncated value by adding 1, and writes bits [16:1] to the destination. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_shuffle_epi8 (__m128i __a, __m128i __b)
 Copies the 8-bit integers from a 128-bit integer vector to the destination or clears 8-bit values in the destination, as specified by the second source operand. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_shuffle_pi8 (__m64 __a, __m64 __b)
 Copies the 8-bit integers from a 64-bit integer vector to the destination or clears 8-bit values in the destination, as specified by the second source operand. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sign_epi8 (__m128i __a, __m128i __b)
 For each 8-bit integer in the first source operand, perform one of the following actions as specified by the second source operand. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sign_epi16 (__m128i __a, __m128i __b)
 For each 16-bit integer in the first source operand, perform one of the following actions as specified by the second source operand. More...
 
static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sign_epi32 (__m128i __a, __m128i __b)
 For each 32-bit integer in the first source operand, perform one of the following actions as specified by the second source operand. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_sign_pi8 (__m64 __a, __m64 __b)
 For each 8-bit integer in the first source operand, perform one of the following actions as specified by the second source operand. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_sign_pi16 (__m64 __a, __m64 __b)
 For each 16-bit integer in the first source operand, perform one of the following actions as specified by the second source operand. More...
 
static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_sign_pi32 (__m64 __a, __m64 __b)
 For each 32-bit integer in the first source operand, perform one of the following actions as specified by the second source operand. More...
 

Macro Definition Documentation

◆ __DEFAULT_FN_ATTRS

#define __DEFAULT_FN_ATTRS   __attribute__((__always_inline__, __nodebug__, __target__("ssse3"), __min_vector_width__(64)))

Definition at line 20 of file tmmintrin.h.

◆ __DEFAULT_FN_ATTRS_MMX

#define __DEFAULT_FN_ATTRS_MMX   __attribute__((__always_inline__, __nodebug__, __target__("mmx,ssse3"), __min_vector_width__(64)))

Definition at line 21 of file tmmintrin.h.

◆ _mm_alignr_epi8

#define _mm_alignr_epi8 (   a,
  b,
 
)
Value:
((__m128i)__builtin_ia32_palignr128((__v16qi)(__m128i)(a), \
(__v16qi)(__m128i)(b), (n)))

Concatenates the two 128-bit integer vector operands, and right-shifts the result by the number of bytes specified in the immediate operand.

__m128i _mm_alignr_epi8(__m128i a, __m128i b, const int n);

This intrinsic corresponds to the PALIGNR instruction.

Parameters
aA 128-bit vector of [16 x i8] containing one of the source operands.
bA 128-bit vector of [16 x i8] containing one of the source operands.
nAn immediate operand specifying how many bytes to right-shift the result.
Returns
A 128-bit integer vector containing the concatenated right-shifted value.

Definition at line 151 of file tmmintrin.h.

◆ _mm_alignr_pi8

#define _mm_alignr_pi8 (   a,
  b,
 
)    ((__m64)__builtin_ia32_palignr((__v8qi)(__m64)(a), (__v8qi)(__m64)(b), (n)))

Concatenates the two 64-bit integer vector operands, and right-shifts the result by the number of bytes specified in the immediate operand.

__m64 _mm_alignr_pi8(__m64 a, __m64 b, const int n);

This intrinsic corresponds to the PALIGNR instruction.

Parameters
aA 64-bit vector of [8 x i8] containing one of the source operands.
bA 64-bit vector of [8 x i8] containing one of the source operands.
nAn immediate operand specifying how many bytes to right-shift the result.
Returns
A 64-bit integer vector containing the concatenated right-shifted value.

Definition at line 174 of file tmmintrin.h.

Function Documentation

◆ _mm_abs_epi16()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_abs_epi16 ( __m128i  __a)
static

Computes the absolute value of each of the packed 16-bit signed integers in the source operand and stores the 16-bit unsigned integer results in the destination.

This intrinsic corresponds to the VPABSW instruction.

Parameters
__aA 128-bit vector of [8 x i16].
Returns
A 128-bit integer vector containing the absolute values of the elements in the operand.

Definition at line 90 of file tmmintrin.h.

References __a.

Referenced by _mm_mask_abs_epi16(), and _mm_maskz_abs_epi16().

◆ _mm_abs_epi32()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_abs_epi32 ( __m128i  __a)
static

Computes the absolute value of each of the packed 32-bit signed integers in the source operand and stores the 32-bit unsigned integer results in the destination.

This intrinsic corresponds to the VPABSD instruction.

Parameters
__aA 128-bit vector of [4 x i32].
Returns
A 128-bit integer vector containing the absolute values of the elements in the operand.

Definition at line 126 of file tmmintrin.h.

References __a.

Referenced by _mm_mask_abs_epi32(), and _mm_maskz_abs_epi32().

◆ _mm_abs_epi8()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_abs_epi8 ( __m128i  __a)
static

Computes the absolute value of each of the packed 8-bit signed integers in the source operand and stores the 8-bit unsigned integer results in the destination.

This intrinsic corresponds to the VPABSB instruction.

Parameters
__aA 128-bit vector of [16 x i8].
Returns
A 128-bit integer vector containing the absolute values of the elements in the operand.

Definition at line 54 of file tmmintrin.h.

References __a.

Referenced by _mm_mask_abs_epi8(), and _mm_maskz_abs_epi8().

◆ _mm_abs_pi16()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_abs_pi16 ( __m64  __a)
static

Computes the absolute value of each of the packed 16-bit signed integers in the source operand and stores the 16-bit unsigned integer results in the destination.

This intrinsic corresponds to the PABSW instruction.

Parameters
__aA 64-bit vector of [4 x i16].
Returns
A 64-bit integer vector containing the absolute values of the elements in the operand.

Definition at line 72 of file tmmintrin.h.

References __a.

◆ _mm_abs_pi32()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_abs_pi32 ( __m64  __a)
static

Computes the absolute value of each of the packed 32-bit signed integers in the source operand and stores the 32-bit unsigned integer results in the destination.

This intrinsic corresponds to the PABSD instruction.

Parameters
__aA 64-bit vector of [2 x i32].
Returns
A 64-bit integer vector containing the absolute values of the elements in the operand.

Definition at line 108 of file tmmintrin.h.

References __a.

◆ _mm_abs_pi8()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_abs_pi8 ( __m64  __a)
static

Computes the absolute value of each of the packed 8-bit signed integers in the source operand and stores the 8-bit unsigned integer results in the destination.

This intrinsic corresponds to the PABSB instruction.

Parameters
__aA 64-bit vector of [8 x i8].
Returns
A 64-bit integer vector containing the absolute values of the elements in the operand.

Definition at line 36 of file tmmintrin.h.

References __a.

◆ _mm_hadd_epi16()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_hadd_epi16 ( __m128i  __a,
__m128i  __b 
)
static

Horizontally adds the adjacent pairs of values contained in 2 packed 128-bit vectors of [8 x i16].

This intrinsic corresponds to the VPHADDW instruction.

Parameters
__aA 128-bit vector of [8 x i16] containing one of the source operands. The horizontal sums of the values are stored in the lower bits of the destination.
__bA 128-bit vector of [8 x i16] containing one of the source operands. The horizontal sums of the values are stored in the upper bits of the destination.
Returns
A 128-bit vector of [8 x i16] containing the horizontal sums of both operands.

Definition at line 195 of file tmmintrin.h.

References __a, and __b.

◆ _mm_hadd_epi32()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_hadd_epi32 ( __m128i  __a,
__m128i  __b 
)
static

Horizontally adds the adjacent pairs of values contained in 2 packed 128-bit vectors of [4 x i32].

This intrinsic corresponds to the VPHADDD instruction.

Parameters
__aA 128-bit vector of [4 x i32] containing one of the source operands. The horizontal sums of the values are stored in the lower bits of the destination.
__bA 128-bit vector of [4 x i32] containing one of the source operands. The horizontal sums of the values are stored in the upper bits of the destination.
Returns
A 128-bit vector of [4 x i32] containing the horizontal sums of both operands.

Definition at line 218 of file tmmintrin.h.

References __a, and __b.

◆ _mm_hadd_pi16()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_hadd_pi16 ( __m64  __a,
__m64  __b 
)
static

Horizontally adds the adjacent pairs of values contained in 2 packed 64-bit vectors of [4 x i16].

This intrinsic corresponds to the PHADDW instruction.

Parameters
__aA 64-bit vector of [4 x i16] containing one of the source operands. The horizontal sums of the values are stored in the lower bits of the destination.
__bA 64-bit vector of [4 x i16] containing one of the source operands. The horizontal sums of the values are stored in the upper bits of the destination.
Returns
A 64-bit vector of [4 x i16] containing the horizontal sums of both operands.

Definition at line 241 of file tmmintrin.h.

References __a, and __b.

◆ _mm_hadd_pi32()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_hadd_pi32 ( __m64  __a,
__m64  __b 
)
static

Horizontally adds the adjacent pairs of values contained in 2 packed 64-bit vectors of [2 x i32].

This intrinsic corresponds to the PHADDD instruction.

Parameters
__aA 64-bit vector of [2 x i32] containing one of the source operands. The horizontal sums of the values are stored in the lower bits of the destination.
__bA 64-bit vector of [2 x i32] containing one of the source operands. The horizontal sums of the values are stored in the upper bits of the destination.
Returns
A 64-bit vector of [2 x i32] containing the horizontal sums of both operands.

Definition at line 264 of file tmmintrin.h.

References __a, and __b.

◆ _mm_hadds_epi16()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_hadds_epi16 ( __m128i  __a,
__m128i  __b 
)
static

Horizontally adds the adjacent pairs of values contained in 2 packed 128-bit vectors of [8 x i16].

Positive sums greater than 0x7FFF are saturated to 0x7FFF. Negative sums less than 0x8000 are saturated to 0x8000.

This intrinsic corresponds to the VPHADDSW instruction.

Parameters
__aA 128-bit vector of [8 x i16] containing one of the source operands. The horizontal sums of the values are stored in the lower bits of the destination.
__bA 128-bit vector of [8 x i16] containing one of the source operands. The horizontal sums of the values are stored in the upper bits of the destination.
Returns
A 128-bit vector of [8 x i16] containing the horizontal saturated sums of both operands.

Definition at line 289 of file tmmintrin.h.

References __a, and __b.

◆ _mm_hadds_pi16()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_hadds_pi16 ( __m64  __a,
__m64  __b 
)
static

Horizontally adds the adjacent pairs of values contained in 2 packed 64-bit vectors of [4 x i16].

Positive sums greater than 0x7FFF are saturated to 0x7FFF. Negative sums less than 0x8000 are saturated to 0x8000.

This intrinsic corresponds to the PHADDSW instruction.

Parameters
__aA 64-bit vector of [4 x i16] containing one of the source operands. The horizontal sums of the values are stored in the lower bits of the destination.
__bA 64-bit vector of [4 x i16] containing one of the source operands. The horizontal sums of the values are stored in the upper bits of the destination.
Returns
A 64-bit vector of [4 x i16] containing the horizontal saturated sums of both operands.

Definition at line 314 of file tmmintrin.h.

References __a, and __b.

◆ _mm_hsub_epi16()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_hsub_epi16 ( __m128i  __a,
__m128i  __b 
)
static

Horizontally subtracts the adjacent pairs of values contained in 2 packed 128-bit vectors of [8 x i16].

This intrinsic corresponds to the VPHSUBW instruction.

Parameters
__aA 128-bit vector of [8 x i16] containing one of the source operands. The horizontal differences between the values are stored in the lower bits of the destination.
__bA 128-bit vector of [8 x i16] containing one of the source operands. The horizontal differences between the values are stored in the upper bits of the destination.
Returns
A 128-bit vector of [8 x i16] containing the horizontal differences of both operands.

Definition at line 337 of file tmmintrin.h.

References __a, and __b.

◆ _mm_hsub_epi32()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_hsub_epi32 ( __m128i  __a,
__m128i  __b 
)
static

Horizontally subtracts the adjacent pairs of values contained in 2 packed 128-bit vectors of [4 x i32].

This intrinsic corresponds to the VPHSUBD instruction.

Parameters
__aA 128-bit vector of [4 x i32] containing one of the source operands. The horizontal differences between the values are stored in the lower bits of the destination.
__bA 128-bit vector of [4 x i32] containing one of the source operands. The horizontal differences between the values are stored in the upper bits of the destination.
Returns
A 128-bit vector of [4 x i32] containing the horizontal differences of both operands.

Definition at line 360 of file tmmintrin.h.

References __a, and __b.

◆ _mm_hsub_pi16()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_hsub_pi16 ( __m64  __a,
__m64  __b 
)
static

Horizontally subtracts the adjacent pairs of values contained in 2 packed 64-bit vectors of [4 x i16].

This intrinsic corresponds to the PHSUBW instruction.

Parameters
__aA 64-bit vector of [4 x i16] containing one of the source operands. The horizontal differences between the values are stored in the lower bits of the destination.
__bA 64-bit vector of [4 x i16] containing one of the source operands. The horizontal differences between the values are stored in the upper bits of the destination.
Returns
A 64-bit vector of [4 x i16] containing the horizontal differences of both operands.

Definition at line 383 of file tmmintrin.h.

References __a, and __b.

◆ _mm_hsub_pi32()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_hsub_pi32 ( __m64  __a,
__m64  __b 
)
static

Horizontally subtracts the adjacent pairs of values contained in 2 packed 64-bit vectors of [2 x i32].

This intrinsic corresponds to the PHSUBD instruction.

Parameters
__aA 64-bit vector of [2 x i32] containing one of the source operands. The horizontal differences between the values are stored in the lower bits of the destination.
__bA 64-bit vector of [2 x i32] containing one of the source operands. The horizontal differences between the values are stored in the upper bits of the destination.
Returns
A 64-bit vector of [2 x i32] containing the horizontal differences of both operands.

Definition at line 406 of file tmmintrin.h.

References __a, and __b.

◆ _mm_hsubs_epi16()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_hsubs_epi16 ( __m128i  __a,
__m128i  __b 
)
static

Horizontally subtracts the adjacent pairs of values contained in 2 packed 128-bit vectors of [8 x i16].

Positive differences greater than 0x7FFF are saturated to 0x7FFF. Negative differences less than 0x8000 are saturated to 0x8000.

This intrinsic corresponds to the VPHSUBSW instruction.

Parameters
__aA 128-bit vector of [8 x i16] containing one of the source operands. The horizontal differences between the values are stored in the lower bits of the destination.
__bA 128-bit vector of [8 x i16] containing one of the source operands. The horizontal differences between the values are stored in the upper bits of the destination.
Returns
A 128-bit vector of [8 x i16] containing the horizontal saturated differences of both operands.

Definition at line 431 of file tmmintrin.h.

References __a, and __b.

◆ _mm_hsubs_pi16()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_hsubs_pi16 ( __m64  __a,
__m64  __b 
)
static

Horizontally subtracts the adjacent pairs of values contained in 2 packed 64-bit vectors of [4 x i16].

Positive differences greater than 0x7FFF are saturated to 0x7FFF. Negative differences less than 0x8000 are saturated to 0x8000.

This intrinsic corresponds to the PHSUBSW instruction.

Parameters
__aA 64-bit vector of [4 x i16] containing one of the source operands. The horizontal differences between the values are stored in the lower bits of the destination.
__bA 64-bit vector of [4 x i16] containing one of the source operands. The horizontal differences between the values are stored in the upper bits of the destination.
Returns
A 64-bit vector of [4 x i16] containing the horizontal saturated differences of both operands.

Definition at line 456 of file tmmintrin.h.

References __a, and __b.

◆ _mm_maddubs_epi16()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_maddubs_epi16 ( __m128i  __a,
__m128i  __b 
)
static

Multiplies corresponding pairs of packed 8-bit unsigned integer values contained in the first source operand and packed 8-bit signed integer values contained in the second source operand, adds pairs of contiguous products with signed saturation, and writes the 16-bit sums to the corresponding bits in the destination.

For example, bits [7:0] of both operands are multiplied, bits [15:8] of both operands are multiplied, and the sum of both results is written to bits [15:0] of the destination.

This intrinsic corresponds to the VPMADDUBSW instruction.

Parameters
__aA 128-bit integer vector containing the first source operand.
__bA 128-bit integer vector containing the second source operand.
Returns
A 128-bit integer vector containing the sums of products of both operands:
R0 := (__a0 * __b0) + (__a1 * __b1)
R1 := (__a2 * __b2) + (__a3 * __b3)
R2 := (__a4 * __b4) + (__a5 * __b5)
R3 := (__a6 * __b6) + (__a7 * __b7)
R4 := (__a8 * __b8) + (__a9 * __b9)
R5 := (__a10 * __b10) + (__a11 * __b11)
R6 := (__a12 * __b12) + (__a13 * __b13)
R7 := (__a14 * __b14) + (__a15 * __b15)

Definition at line 490 of file tmmintrin.h.

References __a, and __b.

Referenced by _mm_mask_maddubs_epi16(), and _mm_maskz_maddubs_epi16().

◆ _mm_maddubs_pi16()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_maddubs_pi16 ( __m64  __a,
__m64  __b 
)
static

Multiplies corresponding pairs of packed 8-bit unsigned integer values contained in the first source operand and packed 8-bit signed integer values contained in the second source operand, adds pairs of contiguous products with signed saturation, and writes the 16-bit sums to the corresponding bits in the destination.

For example, bits [7:0] of both operands are multiplied, bits [15:8] of both operands are multiplied, and the sum of both results is written to bits [15:0] of the destination.

This intrinsic corresponds to the PMADDUBSW instruction.

Parameters
__aA 64-bit integer vector containing the first source operand.
__bA 64-bit integer vector containing the second source operand.
Returns
A 64-bit integer vector containing the sums of products of both operands:
R0 := (__a0 * __b0) + (__a1 * __b1)
R1 := (__a2 * __b2) + (__a3 * __b3)
R2 := (__a4 * __b4) + (__a5 * __b5)
R3 := (__a6 * __b6) + (__a7 * __b7)

Definition at line 520 of file tmmintrin.h.

References __a, and __b.

◆ _mm_mulhrs_epi16()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_mulhrs_epi16 ( __m128i  __a,
__m128i  __b 
)
static

Multiplies packed 16-bit signed integer values, truncates the 32-bit products to the 18 most significant bits by right-shifting, rounds the truncated value by adding 1, and writes bits [16:1] to the destination.

This intrinsic corresponds to the VPMULHRSW instruction.

Parameters
__aA 128-bit vector of [8 x i16] containing one of the source operands.
__bA 128-bit vector of [8 x i16] containing one of the source operands.
Returns
A 128-bit vector of [8 x i16] containing the rounded and scaled products of both operands.

Definition at line 540 of file tmmintrin.h.

References __a, and __b.

Referenced by _mm_mask_mulhrs_epi16(), and _mm_maskz_mulhrs_epi16().

◆ _mm_mulhrs_pi16()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_mulhrs_pi16 ( __m64  __a,
__m64  __b 
)
static

Multiplies packed 16-bit signed integer values, truncates the 32-bit products to the 18 most significant bits by right-shifting, rounds the truncated value by adding 1, and writes bits [16:1] to the destination.

This intrinsic corresponds to the PMULHRSW instruction.

Parameters
__aA 64-bit vector of [4 x i16] containing one of the source operands.
__bA 64-bit vector of [4 x i16] containing one of the source operands.
Returns
A 64-bit vector of [4 x i16] containing the rounded and scaled products of both operands.

Definition at line 560 of file tmmintrin.h.

References __a, and __b.

◆ _mm_shuffle_epi8()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_shuffle_epi8 ( __m128i  __a,
__m128i  __b 
)
static

Copies the 8-bit integers from a 128-bit integer vector to the destination or clears 8-bit values in the destination, as specified by the second source operand.

This intrinsic corresponds to the VPSHUFB instruction.

Parameters
__aA 128-bit integer vector containing the values to be copied.
__bA 128-bit integer vector containing control bytes corresponding to positions in the destination: Bit 7:
1: Clear the corresponding byte in the destination.
0: Copy the selected source byte to the corresponding byte in the destination.
Bits [6:4] Reserved.
Bits [3:0] select the source byte to be copied.
Returns
A 128-bit integer vector containing the copied or cleared values.

Definition at line 586 of file tmmintrin.h.

References __a, and __b.

Referenced by _mm_mask_shuffle_epi8(), and _mm_maskz_shuffle_epi8().

◆ _mm_shuffle_pi8()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_shuffle_pi8 ( __m64  __a,
__m64  __b 
)
static

Copies the 8-bit integers from a 64-bit integer vector to the destination or clears 8-bit values in the destination, as specified by the second source operand.

This intrinsic corresponds to the PSHUFB instruction.

Parameters
__aA 64-bit integer vector containing the values to be copied.
__bA 64-bit integer vector containing control bytes corresponding to positions in the destination: Bit 7:
1: Clear the corresponding byte in the destination.
0: Copy the selected source byte to the corresponding byte in the destination.
Bits [3:0] select the source byte to be copied.
Returns
A 64-bit integer vector containing the copied or cleared values.

Definition at line 611 of file tmmintrin.h.

References __a, and __b.

◆ _mm_sign_epi16()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sign_epi16 ( __m128i  __a,
__m128i  __b 
)
static

For each 16-bit integer in the first source operand, perform one of the following actions as specified by the second source operand.

If the word in the second source is negative, calculate the two's complement of the corresponding word in the first source, and write that value to the destination. If the word in the second source is positive, copy the corresponding word from the first source to the destination. If the word in the second source is zero, clear the corresponding word in the destination.

This intrinsic corresponds to the VPSIGNW instruction.

Parameters
__aA 128-bit integer vector containing the values to be copied.
__bA 128-bit integer vector containing control words corresponding to positions in the destination.
Returns
A 128-bit integer vector containing the resultant values.

Definition at line 663 of file tmmintrin.h.

References __a, and __b.

◆ _mm_sign_epi32()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sign_epi32 ( __m128i  __a,
__m128i  __b 
)
static

For each 32-bit integer in the first source operand, perform one of the following actions as specified by the second source operand.

If the doubleword in the second source is negative, calculate the two's complement of the corresponding word in the first source, and write that value to the destination. If the doubleword in the second source is positive, copy the corresponding word from the first source to the destination. If the doubleword in the second source is zero, clear the corresponding word in the destination.

This intrinsic corresponds to the VPSIGND instruction.

Parameters
__aA 128-bit integer vector containing the values to be copied.
__bA 128-bit integer vector containing control doublewords corresponding to positions in the destination.
Returns
A 128-bit integer vector containing the resultant values.

Definition at line 689 of file tmmintrin.h.

References __a, and __b.

◆ _mm_sign_epi8()

static __inline__ __m128i __DEFAULT_FN_ATTRS _mm_sign_epi8 ( __m128i  __a,
__m128i  __b 
)
static

For each 8-bit integer in the first source operand, perform one of the following actions as specified by the second source operand.

If the byte in the second source is negative, calculate the two's complement of the corresponding byte in the first source, and write that value to the destination. If the byte in the second source is positive, copy the corresponding byte from the first source to the destination. If the byte in the second source is zero, clear the corresponding byte in the destination.

This intrinsic corresponds to the VPSIGNB instruction.

Parameters
__aA 128-bit integer vector containing the values to be copied.
__bA 128-bit integer vector containing control bytes corresponding to positions in the destination.
Returns
A 128-bit integer vector containing the resultant values.

Definition at line 637 of file tmmintrin.h.

References __a, and __b.

◆ _mm_sign_pi16()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_sign_pi16 ( __m64  __a,
__m64  __b 
)
static

For each 16-bit integer in the first source operand, perform one of the following actions as specified by the second source operand.

If the word in the second source is negative, calculate the two's complement of the corresponding word in the first source, and write that value to the destination. If the word in the second source is positive, copy the corresponding word from the first source to the destination. If the word in the second source is zero, clear the corresponding word in the destination.

This intrinsic corresponds to the PSIGNW instruction.

Parameters
__aA 64-bit integer vector containing the values to be copied.
__bA 64-bit integer vector containing control words corresponding to positions in the destination.
Returns
A 64-bit integer vector containing the resultant values.

Definition at line 741 of file tmmintrin.h.

References __a, and __b.

◆ _mm_sign_pi32()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_sign_pi32 ( __m64  __a,
__m64  __b 
)
static

For each 32-bit integer in the first source operand, perform one of the following actions as specified by the second source operand.

If the doubleword in the second source is negative, calculate the two's complement of the corresponding doubleword in the first source, and write that value to the destination. If the doubleword in the second source is positive, copy the corresponding doubleword from the first source to the destination. If the doubleword in the second source is zero, clear the corresponding doubleword in the destination.

This intrinsic corresponds to the PSIGND instruction.

Parameters
__aA 64-bit integer vector containing the values to be copied.
__bA 64-bit integer vector containing two control doublewords corresponding to positions in the destination.
Returns
A 64-bit integer vector containing the resultant values.

Definition at line 767 of file tmmintrin.h.

References __a, and __b.

◆ _mm_sign_pi8()

static __inline__ __m64 __DEFAULT_FN_ATTRS_MMX _mm_sign_pi8 ( __m64  __a,
__m64  __b 
)
static

For each 8-bit integer in the first source operand, perform one of the following actions as specified by the second source operand.

If the byte in the second source is negative, calculate the two's complement of the corresponding byte in the first source, and write that value to the destination. If the byte in the second source is positive, copy the corresponding byte from the first source to the destination. If the byte in the second source is zero, clear the corresponding byte in the destination.

This intrinsic corresponds to the PSIGNB instruction.

Parameters
__aA 64-bit integer vector containing the values to be copied.
__bA 64-bit integer vector containing control bytes corresponding to positions in the destination.
Returns
A 64-bit integer vector containing the resultant values.

Definition at line 715 of file tmmintrin.h.

References __a, and __b.

b
__device__ __2f16 b
Definition: __clang_hip_libdevice_declares.h:314
_mm_alignr_epi8
#define _mm_alignr_epi8(a, b, n)
Concatenates the two 128-bit integer vector operands, and right-shifts the result by the number of by...
Definition: tmmintrin.h:151
_mm_alignr_pi8
#define _mm_alignr_pi8(a, b, n)
Concatenates the two 64-bit integer vector operands, and right-shifts the result by the number of byt...
Definition: tmmintrin.h:174