src/simd/archsimd/ops_internal_amd64.go

// Code generated by x/arch/internal/simdgen using 'go run . -xedPath $XED_PATH -o godefs -goroot $GOROOT go.yaml types.yaml categories.yaml'; DO NOT EDIT.

//go:build goexperiment.simd

package archsimd

/* blend */

// blend blends two vectors based on mask values, choosing either
// the first or the second based on whether the third is false or true
//
// Asm: VPBLENDVB, CPU Feature: AVX
func (x Int8x16) blend(y Int8x16, mask Int8x16) Int8x16

// blend blends two vectors based on mask values, choosing either
// the first or the second based on whether the third is false or true
//
// Asm: VPBLENDVB, CPU Feature: AVX2
func (x Int8x32) blend(y Int8x32, mask Int8x32) Int8x32

/* blendMasked */

// blendMasked blends two vectors based on mask values, choosing either
// the first or the second based on whether the third is false or true
//
// This operation is applied selectively under a write mask.
//
// Asm: VPBLENDMB, CPU Feature: AVX512
func (x Int8x64) blendMasked(y Int8x64, mask Mask8x64) Int8x64

// blendMasked blends two vectors based on mask values, choosing either
// the first or the second based on whether the third is false or true
//
// This operation is applied selectively under a write mask.
//
// Asm: VPBLENDMW, CPU Feature: AVX512
func (x Int16x32) blendMasked(y Int16x32, mask Mask16x32) Int16x32

// blendMasked blends two vectors based on mask values, choosing either
// the first or the second based on whether the third is false or true
//
// This operation is applied selectively under a write mask.
//
// Asm: VPBLENDMD, CPU Feature: AVX512
func (x Int32x16) blendMasked(y Int32x16, mask Mask32x16) Int32x16

// blendMasked blends two vectors based on mask values, choosing either
// the first or the second based on whether the third is false or true
//
// This operation is applied selectively under a write mask.
//
// Asm: VPBLENDMQ, CPU Feature: AVX512
func (x Int64x8) blendMasked(y Int64x8, mask Mask64x8) Int64x8

/* carrylessMultiply */

// carrylessMultiply computes one of four possible Galois polynomial
// products of selected high and low halves of x and y,
// depending on the value of xyHiLo, returning the 128-bit
// product in the concatenated two elements of the result.
// Bit 0 selects the low (0) or high (1) element of x and
// bit 4 selects the low (0x00) or high (0x10) element of y.
//
// xyHiLo results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPCLMULQDQ, CPU Feature: AVX
func (x Uint64x2) carrylessMultiply(xyHiLo uint8, y Uint64x2) Uint64x2

// carrylessMultiply computes one of two possible Galois polynomial
// products of selected high and low halves of each of the two
// 128-bit lanes of x and y, depending on the value of xyHiLo,
// and returns the four 128-bit products in the result's lanes.
// Bit 0 selects the low (0) or high (1) elements of x's lanes and
// bit 4 selects the low (0x00) or high (0x10) elements of y's lanes.
//
// xyHiLo results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPCLMULQDQ, CPU Feature: AVX512VPCLMULQDQ
func (x Uint64x4) carrylessMultiply(xyHiLo uint8, y Uint64x4) Uint64x4

// carrylessMultiply computes one of four possible Galois polynomial
// products of selected high and low halves of each of the four
// 128-bit lanes of x and y, depending on the value of xyHiLo,
// and returns the four 128-bit products in the result's lanes.
// Bit 0 selects the low (0) or high (1) elements of x's lanes and
// bit 4 selects the low (0x00) or high (0x10) elements of y's lanes.
//
// xyHiLo results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPCLMULQDQ, CPU Feature: AVX512VPCLMULQDQ
func (x Uint64x8) carrylessMultiply(xyHiLo uint8, y Uint64x8) Uint64x8

/* concatSelectedConstant */

// concatSelectedConstant concatenates selected elements from x and y into the lower and upper
// halves of the output.  The selection is chosen by the constant parameter h1h0l1l0
// where each {h,l}{1,0} is two bits specify which element from y or x to select.
// For example, {0,1,2,3}.concatSelectedConstant(0b_11_01_00_10, {4,5,6,7}) returns
// {2, 0, 5, 7} (don't forget that the binary constant is written big-endian).
//
// h1h0l1l0 results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPS, CPU Feature: AVX
func (x Float32x4) concatSelectedConstant(h1h0l1l0 uint8, y Float32x4) Float32x4

// concatSelectedConstant concatenates selected elements from x and y into the lower and upper
// halves of the output.  The selection is chosen by the constant parameter hilo
// where hi and lo are each one bit specifying which 64-bit element to select
// from y and x.  For example {4,5}.concatSelectedConstant(0b10, {6,7})
// returns {4,7}; bit 0, selecting from x, is zero, and selects 4, and bit 1,
// selecting from y, is 1, and selects 7.
//
// hilo results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPD, CPU Feature: AVX
func (x Float64x2) concatSelectedConstant(hilo uint8, y Float64x2) Float64x2

// concatSelectedConstant concatenates selected elements from x and y into the lower and upper
// halves of the output.  The selection is chosen by the constant parameter h1h0l1l0
// where each {h,l}{1,0} is two bits specify which element from y or x to select.
// For example, {0,1,2,3}.concatSelectedConstant(0b_11_01_00_10, {4,5,6,7}) returns
// {2, 0, 5, 7} (don't forget that the binary constant is written big-endian).
//
// h1h0l1l0 results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPS, CPU Feature: AVX
func (x Int32x4) concatSelectedConstant(h1h0l1l0 uint8, y Int32x4) Int32x4

// concatSelectedConstant concatenates selected elements from x and y into the lower and upper
// halves of the output.  The selection is chosen by the constant parameter hilo
// where hi and lo are each one bit specifying which 64-bit element to select
// from y and x.  For example {4,5}.concatSelectedConstant(0b10, {6,7})
// returns {4,7}; bit 0, selecting from x, is zero, and selects 4, and bit 1,
// selecting from y, is 1, and selects 7.
//
// hilo results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPD, CPU Feature: AVX
func (x Int64x2) concatSelectedConstant(hilo uint8, y Int64x2) Int64x2

// concatSelectedConstant concatenates selected elements from x and y into the lower and upper
// halves of the output.  The selection is chosen by the constant parameter h1h0l1l0
// where each {h,l}{1,0} is two bits specify which element from y or x to select.
// For example, {0,1,2,3}.concatSelectedConstant(0b_11_01_00_10, {4,5,6,7}) returns
// {2, 0, 5, 7} (don't forget that the binary constant is written big-endian).
//
// h1h0l1l0 results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPS, CPU Feature: AVX
func (x Uint32x4) concatSelectedConstant(h1h0l1l0 uint8, y Uint32x4) Uint32x4

// concatSelectedConstant concatenates selected elements from x and y into the lower and upper
// halves of the output.  The selection is chosen by the constant parameter hilo
// where hi and lo are each one bit specifying which 64-bit element to select
// from y and x.  For example {4,5}.concatSelectedConstant(0b10, {6,7})
// returns {4,7}; bit 0, selecting from x, is zero, and selects 4, and bit 1,
// selecting from y, is 1, and selects 7.
//
// hilo results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPD, CPU Feature: AVX
func (x Uint64x2) concatSelectedConstant(hilo uint8, y Uint64x2) Uint64x2

/* concatSelectedConstantGrouped */

// concatSelectedConstantGrouped concatenates selected elements from 128-bit subvectors of x and y
// into the lower and upper halves of corresponding subvectors of the output.
// The selection is chosen by the constant parameter h1h0l1l0
// where each {h,l}{1,0} is two bits specifying which element from y or x to select.
// For example,
// {0,1,2,3,8,9,10,11}.concatSelectedConstantGrouped(0b_11_01_00_10, {4,5,6,7,12,13,14,15})
// returns {2,0,5,7,10,8,13,15}
// (don't forget that the binary constant is written big-endian).
//
// h1h0l1l0 results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPS, CPU Feature: AVX
func (x Float32x8) concatSelectedConstantGrouped(h1h0l1l0 uint8, y Float32x8) Float32x8

// concatSelectedConstantGrouped concatenates selected elements from 128-bit subvectors of x and y
// into the lower and upper halves of corresponding subvectors of the output.
// The selection is chosen by the constant parameter h1h0l1l0
// where each {h,l}{1,0} is two bits specifying which element from y or x to select.
// For example,
//
//	{0,1,2,3,8,9,10,11, 20,21,22,23,28,29,210,211}.concatSelectedConstantGrouped(
//	 0b_11_01_00_10, {4,5,6,7,12,13,14,15, 24,25,26,27,212,213,214,215})
//
// returns {2,0,5,7,10,8,13,15, 22,20,25,27,210,28,213,215}
//
// (don't forget that the binary constant is written big-endian).
//
// h1h0l1l0 results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPS, CPU Feature: AVX512
func (x Float32x16) concatSelectedConstantGrouped(h1h0l1l0 uint8, y Float32x16) Float32x16

// concatSelectedConstantGrouped concatenates selected elements from 128-bit subvectors of x and y
// into the lower and upper halves of corresponding subvectors of the output.
// The selections are specified by the constant parameter hilos where each
// hi and lo pair select 64-bit elements from the corresponding 128-bit
// subvectors of x and y.
//
// For example {4,5,8,9}.concatSelectedConstantGrouped(0b_11_10, {6,7,10,11})
// returns {4,7,9,11}; bit 0 is zero, selecting element 0 from x's least
// 128-bits (4), then 1, selects the element 1 from y's least 128-bits (7),
// then 1, selecting element 1 from x's upper 128 bits (9), then 1,
// selecting element 1 from y's upper 128 bits (11).
// This differs from the same method applied to a 32x8 vector, where
// the 8-bit constant performs the same selection on both subvectors.
//
// hilos results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPD, CPU Feature: AVX
func (x Float64x4) concatSelectedConstantGrouped(hilos uint8, y Float64x4) Float64x4

// concatSelectedConstantGrouped concatenates selected elements from 128-bit subvectors of x and y
// into the lower and upper halves of corresponding subvectors of the output.
// The selections are specified by the constant parameter hilos where each
// hi and lo pair select 64-bit elements from the corresponding 128-bit
// subvectors of x and y.
//
// For example {4,5,8,9,12,13,16,17}.concatSelectedConstantGrouped(0b11_00_11_10, {6,7,10,11,14,15,18,19})
// returns {4,7,9,11,12,14,17,19}; bit 0 is zero, selecting element 0 from x's
// least 128-bits (4), then 1, selects the element 1 from y's least 128-bits (7),
// then 1, selecting element 1 from x's next 128 bits (9), then 1,
// selecting element 1 from y's upper 128 bits (11).  The next two 0 bits select
// the lower elements from x and y's 3rd 128 bit groups (12, 14), the last two
// 1 bits select the upper elements from x and y's last 128 bits (17, 19).
// This differs from the same method applied to a 32x8 or 32x16 vector, where
// the 8-bit constant performs the same selection on all the subvectors.
//
// hilos results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPD, CPU Feature: AVX512
func (x Float64x8) concatSelectedConstantGrouped(hilos uint8, y Float64x8) Float64x8

// concatSelectedConstantGrouped concatenates selected elements from 128-bit subvectors of x and y
// into the lower and upper halves of corresponding subvectors of the output.
// The selection is chosen by the constant parameter h1h0l1l0
// where each {h,l}{1,0} is two bits specifying which element from y or x to select.
// For example,
// {0,1,2,3,8,9,10,11}.concatSelectedConstantGrouped(0b_11_01_00_10, {4,5,6,7,12,13,14,15})
// returns {2,0,5,7,10,8,13,15}
// (don't forget that the binary constant is written big-endian).
//
// h1h0l1l0 results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPS, CPU Feature: AVX
func (x Int32x8) concatSelectedConstantGrouped(h1h0l1l0 uint8, y Int32x8) Int32x8

// concatSelectedConstantGrouped concatenates selected elements from 128-bit subvectors of x and y
// into the lower and upper halves of corresponding subvectors of the output.
// The selection is chosen by the constant parameter h1h0l1l0
// where each {h,l}{1,0} is two bits specifying which element from y or x to select.
// For example,
//
//	{0,1,2,3,8,9,10,11, 20,21,22,23,28,29,210,211}.concatSelectedConstantGrouped(
//	 0b_11_01_00_10, {4,5,6,7,12,13,14,15, 24,25,26,27,212,213,214,215})
//
// returns {2,0,5,7,10,8,13,15, 22,20,25,27,210,28,213,215}
//
// (don't forget that the binary constant is written big-endian).
//
// h1h0l1l0 results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPS, CPU Feature: AVX512
func (x Int32x16) concatSelectedConstantGrouped(h1h0l1l0 uint8, y Int32x16) Int32x16

// concatSelectedConstantGrouped concatenates selected elements from 128-bit subvectors of x and y
// into the lower and upper halves of corresponding subvectors of the output.
// The selections are specified by the constant parameter hilos where each
// hi and lo pair select 64-bit elements from the corresponding 128-bit
// subvectors of x and y.
//
// For example {4,5,8,9}.concatSelectedConstantGrouped(0b_11_10, {6,7,10,11})
// returns {4,7,9,11}; bit 0 is zero, selecting element 0 from x's least
// 128-bits (4), then 1, selects the element 1 from y's least 128-bits (7),
// then 1, selecting element 1 from x's upper 128 bits (9), then 1,
// selecting element 1 from y's upper 128 bits (11).
// This differs from the same method applied to a 32x8 vector, where
// the 8-bit constant performs the same selection on both subvectors.
//
// hilos results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPD, CPU Feature: AVX
func (x Int64x4) concatSelectedConstantGrouped(hilos uint8, y Int64x4) Int64x4

// concatSelectedConstantGrouped concatenates selected elements from 128-bit subvectors of x and y
// into the lower and upper halves of corresponding subvectors of the output.
// The selections are specified by the constant parameter hilos where each
// hi and lo pair select 64-bit elements from the corresponding 128-bit
// subvectors of x and y.
//
// For example {4,5,8,9,12,13,16,17}.concatSelectedConstantGrouped(0b11_00_11_10, {6,7,10,11,14,15,18,19})
// returns {4,7,9,11,12,14,17,19}; bit 0 is zero, selecting element 0 from x's
// least 128-bits (4), then 1, selects the element 1 from y's least 128-bits (7),
// then 1, selecting element 1 from x's next 128 bits (9), then 1,
// selecting element 1 from y's upper 128 bits (11).  The next two 0 bits select
// the lower elements from x and y's 3rd 128 bit groups (12, 14), the last two
// 1 bits select the upper elements from x and y's last 128 bits (17, 19).
// This differs from the same method applied to a 32x8 or 32x16 vector, where
// the 8-bit constant performs the same selection on all the subvectors.
//
// hilos results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPD, CPU Feature: AVX512
func (x Int64x8) concatSelectedConstantGrouped(hilos uint8, y Int64x8) Int64x8

// concatSelectedConstantGrouped concatenates selected elements from 128-bit subvectors of x and y
// into the lower and upper halves of corresponding subvectors of the output.
// The selection is chosen by the constant parameter h1h0l1l0
// where each {h,l}{1,0} is two bits specifying which element from y or x to select.
// For example,
// {0,1,2,3,8,9,10,11}.concatSelectedConstantGrouped(0b_11_01_00_10, {4,5,6,7,12,13,14,15})
// returns {2,0,5,7,10,8,13,15}
// (don't forget that the binary constant is written big-endian).
//
// h1h0l1l0 results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPS, CPU Feature: AVX
func (x Uint32x8) concatSelectedConstantGrouped(h1h0l1l0 uint8, y Uint32x8) Uint32x8

// concatSelectedConstantGrouped concatenates selected elements from 128-bit subvectors of x and y
// into the lower and upper halves of corresponding subvectors of the output.
// The selection is chosen by the constant parameter h1h0l1l0
// where each {h,l}{1,0} is two bits specifying which element from y or x to select.
// For example,
//
//	{0,1,2,3,8,9,10,11, 20,21,22,23,28,29,210,211}.concatSelectedConstantGrouped(
//	 0b_11_01_00_10, {4,5,6,7,12,13,14,15, 24,25,26,27,212,213,214,215})
//
// returns {2,0,5,7,10,8,13,15, 22,20,25,27,210,28,213,215}
//
// (don't forget that the binary constant is written big-endian).
//
// h1h0l1l0 results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPS, CPU Feature: AVX512
func (x Uint32x16) concatSelectedConstantGrouped(h1h0l1l0 uint8, y Uint32x16) Uint32x16

// concatSelectedConstantGrouped concatenates selected elements from 128-bit subvectors of x and y
// into the lower and upper halves of corresponding subvectors of the output.
// The selections are specified by the constant parameter hilos where each
// hi and lo pair select 64-bit elements from the corresponding 128-bit
// subvectors of x and y.
//
// For example {4,5,8,9}.concatSelectedConstantGrouped(0b_11_10, {6,7,10,11})
// returns {4,7,9,11}; bit 0 is zero, selecting element 0 from x's least
// 128-bits (4), then 1, selects the element 1 from y's least 128-bits (7),
// then 1, selecting element 1 from x's upper 128 bits (9), then 1,
// selecting element 1 from y's upper 128 bits (11).
// This differs from the same method applied to a 32x8 vector, where
// the 8-bit constant performs the same selection on both subvectors.
//
// hilos results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPD, CPU Feature: AVX
func (x Uint64x4) concatSelectedConstantGrouped(hilos uint8, y Uint64x4) Uint64x4

// concatSelectedConstantGrouped concatenates selected elements from 128-bit subvectors of x and y
// into the lower and upper halves of corresponding subvectors of the output.
// The selections are specified by the constant parameter hilos where each
// hi and lo pair select 64-bit elements from the corresponding 128-bit
// subvectors of x and y.
//
// For example {4,5,8,9,12,13,16,17}.concatSelectedConstantGrouped(0b11_00_11_10, {6,7,10,11,14,15,18,19})
// returns {4,7,9,11,12,14,17,19}; bit 0 is zero, selecting element 0 from x's
// least 128-bits (4), then 1, selects the element 1 from y's least 128-bits (7),
// then 1, selecting element 1 from x's next 128 bits (9), then 1,
// selecting element 1 from y's upper 128 bits (11).  The next two 0 bits select
// the lower elements from x and y's 3rd 128 bit groups (12, 14), the last two
// 1 bits select the upper elements from x and y's last 128 bits (17, 19).
// This differs from the same method applied to a 32x8 or 32x16 vector, where
// the 8-bit constant performs the same selection on all the subvectors.
//
// hilos results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VSHUFPD, CPU Feature: AVX512
func (x Uint64x8) concatSelectedConstantGrouped(hilos uint8, y Uint64x8) Uint64x8

/* permuteScalars */

// permuteScalars performs a permutation of vector x using constant indices:
// result = {x[indices[0:2]], x[indices[2:4]], x[indices[4:6]], x[indices[6:8]]}
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFD, CPU Feature: AVX
func (x Int32x4) permuteScalars(indices uint8) Int32x4

// permuteScalars performs a permutation of vector x using constant indices:
// result = {x[indices[0:2]], x[indices[2:4]], x[indices[4:6]], x[indices[6:8]]}
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFD, CPU Feature: AVX
func (x Uint32x4) permuteScalars(indices uint8) Uint32x4

/* permuteScalarsGrouped */

// permuteScalarsGrouped performs a grouped permutation of vector x using constant indices:
// result = {x_group0[indices[0:2]], x_group0[indices[2:4]], x_group0[indices[4:6]], x_group0[indices[6:8]], x_group1[indices[0:2]], ...}
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
// Each group is of size 128-bit.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFD, CPU Feature: AVX2
func (x Int32x8) permuteScalarsGrouped(indices uint8) Int32x8

// permuteScalarsGrouped performs a grouped permutation of vector x using constant indices:
// result = {x_group0[indices[0:2]], x_group0[indices[2:4]], x_group0[indices[4:6]], x_group0[indices[6:8]], x_group1[indices[0:2]], ...}
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
// Each group is of size 128-bit.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFD, CPU Feature: AVX512
func (x Int32x16) permuteScalarsGrouped(indices uint8) Int32x16

// permuteScalarsGrouped performs a grouped permutation of vector x using constant indices:
// result = {x_group0[indices[0:2]], x_group0[indices[2:4]], x_group0[indices[4:6]], x_group0[indices[6:8]], x_group1[indices[0:2]], ...}
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
// Each group is of size 128-bit.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFD, CPU Feature: AVX2
func (x Uint32x8) permuteScalarsGrouped(indices uint8) Uint32x8

// permuteScalarsGrouped performs a grouped permutation of vector x using constant indices:
// result = {x_group0[indices[0:2]], x_group0[indices[2:4]], x_group0[indices[4:6]], x_group0[indices[6:8]], x_group1[indices[0:2]], ...}
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
// Each group is of size 128-bit.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFD, CPU Feature: AVX512
func (x Uint32x16) permuteScalarsGrouped(indices uint8) Uint32x16

/* permuteScalarsHi */

// permuteScalarsHi performs a permutation of vector x using constant indices:
// result = {x[0], x[1], x[2], x[3], x[indices[0:2]+4], x[indices[2:4]+4], x[indices[4:6]+4], x[indices[6:8]+4]}
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFHW, CPU Feature: AVX512
func (x Int16x8) permuteScalarsHi(indices uint8) Int16x8

// permuteScalarsHi performs a permutation of vector x using constant indices:
// result = {x[0], x[1], x[2], x[3], x[indices[0:2]+4], x[indices[2:4]+4], x[indices[4:6]+4], x[indices[6:8]+4]}
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFHW, CPU Feature: AVX512
func (x Uint16x8) permuteScalarsHi(indices uint8) Uint16x8

/* permuteScalarsHiGrouped */

// permuteScalarsHiGrouped performs a grouped permutation of vector x using constant indices:
// result =
//
//	{x_group0[0], x_group0[1], x_group0[2], x_group0[3], x_group0[indices[0:2]+4], x_group0[indices[2:4]+4], x_group0[indices[4:6]+4], x_group0[indices[6:8]+4],
//	 x_group1[0], x_group1[1], x_group1[2], x_group1[3], x_group1[indices[0:2]+4], ...}
//
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
// Each group is of size 128-bit.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFHW, CPU Feature: AVX2
func (x Int16x16) permuteScalarsHiGrouped(indices uint8) Int16x16

// permuteScalarsHiGrouped performs a grouped permutation of vector x using constant indices:
// result =
//
//	{x_group0[0], x_group0[1], x_group0[2], x_group0[3], x_group0[indices[0:2]+4], x_group0[indices[2:4]+4], x_group0[indices[4:6]+4], x_group0[indices[6:8]+4],
//	 x_group1[0], x_group1[1], x_group1[2], x_group1[3], x_group1[indices[0:2]+4], ...}
//
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
// Each group is of size 128-bit.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFHW, CPU Feature: AVX512
func (x Int16x32) permuteScalarsHiGrouped(indices uint8) Int16x32

// permuteScalarsHiGrouped performs a grouped permutation of vector x using constant indices:
// result =
//
//	{x_group0[0], x_group0[1], x_group0[2], x_group0[3], x_group0[indices[0:2]+4], x_group0[indices[2:4]+4], x_group0[indices[4:6]+4], x_group0[indices[6:8]+4],
//	 x_group1[0], x_group1[1], x_group1[2], x_group1[3], x_group1[indices[0:2]+4], ...}
//
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
// Each group is of size 128-bit.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFHW, CPU Feature: AVX2
func (x Uint16x16) permuteScalarsHiGrouped(indices uint8) Uint16x16

// permuteScalarsHiGrouped performs a grouped permutation of vector x using constant indices:
// result =
//
//	{x_group0[0], x_group0[1], x_group0[2], x_group0[3], x_group0[indices[0:2]+4], x_group0[indices[2:4]+4], x_group0[indices[4:6]+4], x_group0[indices[6:8]+4],
//	 x_group1[0], x_group1[1], x_group1[2], x_group1[3], x_group1[indices[0:2]+4], ...}
//
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
// Each group is of size 128-bit.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFHW, CPU Feature: AVX512
func (x Uint16x32) permuteScalarsHiGrouped(indices uint8) Uint16x32

/* permuteScalarsLo */

// permuteScalarsLo performs a permutation of vector x using constant indices:
// result = {x[indices[0:2]], x[indices[2:4]], x[indices[4:6]], x[indices[6:8]], x[4], x[5], x[6], x[7]}
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFLW, CPU Feature: AVX512
func (x Int16x8) permuteScalarsLo(indices uint8) Int16x8

// permuteScalarsLo performs a permutation of vector x using constant indices:
// result = {x[indices[0:2]], x[indices[2:4]], x[indices[4:6]], x[indices[6:8]], x[4], x[5], x[6], x[7]}
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFLW, CPU Feature: AVX512
func (x Uint16x8) permuteScalarsLo(indices uint8) Uint16x8

/* permuteScalarsLoGrouped */

// permuteScalarsLoGrouped performs a grouped permutation of vector x using constant indices:
//
//	result = {x_group0[indices[0:2]], x_group0[indices[2:4]], x_group0[indices[4:6]], x_group0[indices[6:8]], x[4], x[5], x[6], x[7],
//	 x_group1[indices[0:2]], ...}
//
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
// Each group is of size 128-bit.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFLW, CPU Feature: AVX2
func (x Int16x16) permuteScalarsLoGrouped(indices uint8) Int16x16

// permuteScalarsLoGrouped performs a grouped permutation of vector x using constant indices:
//
//	result = {x_group0[indices[0:2]], x_group0[indices[2:4]], x_group0[indices[4:6]], x_group0[indices[6:8]], x[4], x[5], x[6], x[7],
//	 x_group1[indices[0:2]], ...}
//
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
// Each group is of size 128-bit.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFLW, CPU Feature: AVX512
func (x Int16x32) permuteScalarsLoGrouped(indices uint8) Int16x32

// permuteScalarsLoGrouped performs a grouped permutation of vector x using constant indices:
//
//	result = {x_group0[indices[0:2]], x_group0[indices[2:4]], x_group0[indices[4:6]], x_group0[indices[6:8]], x[4], x[5], x[6], x[7],
//	 x_group1[indices[0:2]], ...}
//
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
// Each group is of size 128-bit.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFLW, CPU Feature: AVX2
func (x Uint16x16) permuteScalarsLoGrouped(indices uint8) Uint16x16

// permuteScalarsLoGrouped performs a grouped permutation of vector x using constant indices:
//
//	result = {x_group0[indices[0:2]], x_group0[indices[2:4]], x_group0[indices[4:6]], x_group0[indices[6:8]], x[4], x[5], x[6], x[7],
//	 x_group1[indices[0:2]], ...}
//
// Indices is four 2-bit values packed into a byte, thus indices[0:2] is the first index.
// Each group is of size 128-bit.
//
// indices results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPSHUFLW, CPU Feature: AVX512
func (x Uint16x32) permuteScalarsLoGrouped(indices uint8) Uint16x32

/* tern */

// tern performs a logical operation on three vectors based on the 8-bit truth table.
// Bitwise, the result is equal to 1 & (table >> (x<<2 + y<<1 + z))
//
// table results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPTERNLOGD, CPU Feature: AVX512
func (x Int32x4) tern(table uint8, y Int32x4, z Int32x4) Int32x4

// tern performs a logical operation on three vectors based on the 8-bit truth table.
// Bitwise, the result is equal to 1 & (table >> (x<<2 + y<<1 + z))
//
// table results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPTERNLOGD, CPU Feature: AVX512
func (x Int32x8) tern(table uint8, y Int32x8, z Int32x8) Int32x8

// tern performs a logical operation on three vectors based on the 8-bit truth table.
// Bitwise, the result is equal to 1 & (table >> (x<<2 + y<<1 + z))
//
// table results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPTERNLOGD, CPU Feature: AVX512
func (x Int32x16) tern(table uint8, y Int32x16, z Int32x16) Int32x16

// tern performs a logical operation on three vectors based on the 8-bit truth table.
// Bitwise, the result is equal to 1 & (table >> (x<<2 + y<<1 + z))
//
// table results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPTERNLOGQ, CPU Feature: AVX512
func (x Int64x2) tern(table uint8, y Int64x2, z Int64x2) Int64x2

// tern performs a logical operation on three vectors based on the 8-bit truth table.
// Bitwise, the result is equal to 1 & (table >> (x<<2 + y<<1 + z))
//
// table results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPTERNLOGQ, CPU Feature: AVX512
func (x Int64x4) tern(table uint8, y Int64x4, z Int64x4) Int64x4

// tern performs a logical operation on three vectors based on the 8-bit truth table.
// Bitwise, the result is equal to 1 & (table >> (x<<2 + y<<1 + z))
//
// table results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPTERNLOGQ, CPU Feature: AVX512
func (x Int64x8) tern(table uint8, y Int64x8, z Int64x8) Int64x8

// tern performs a logical operation on three vectors based on the 8-bit truth table.
// Bitwise, the result is equal to 1 & (table >> (x<<2 + y<<1 + z))
//
// table results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPTERNLOGD, CPU Feature: AVX512
func (x Uint32x4) tern(table uint8, y Uint32x4, z Uint32x4) Uint32x4

// tern performs a logical operation on three vectors based on the 8-bit truth table.
// Bitwise, the result is equal to 1 & (table >> (x<<2 + y<<1 + z))
//
// table results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPTERNLOGD, CPU Feature: AVX512
func (x Uint32x8) tern(table uint8, y Uint32x8, z Uint32x8) Uint32x8

// tern performs a logical operation on three vectors based on the 8-bit truth table.
// Bitwise, the result is equal to 1 & (table >> (x<<2 + y<<1 + z))
//
// table results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPTERNLOGD, CPU Feature: AVX512
func (x Uint32x16) tern(table uint8, y Uint32x16, z Uint32x16) Uint32x16

// tern performs a logical operation on three vectors based on the 8-bit truth table.
// Bitwise, the result is equal to 1 & (table >> (x<<2 + y<<1 + z))
//
// table results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPTERNLOGQ, CPU Feature: AVX512
func (x Uint64x2) tern(table uint8, y Uint64x2, z Uint64x2) Uint64x2

// tern performs a logical operation on three vectors based on the 8-bit truth table.
// Bitwise, the result is equal to 1 & (table >> (x<<2 + y<<1 + z))
//
// table results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPTERNLOGQ, CPU Feature: AVX512
func (x Uint64x4) tern(table uint8, y Uint64x4, z Uint64x4) Uint64x4

// tern performs a logical operation on three vectors based on the 8-bit truth table.
// Bitwise, the result is equal to 1 & (table >> (x<<2 + y<<1 + z))
//
// table results in better performance when it's a constant, a non-constant value will be translated into a jump table.
//
// Asm: VPTERNLOGQ, CPU Feature: AVX512
func (x Uint64x8) tern(table uint8, y Uint64x8, z Uint64x8) Uint64x8