Struct Arch 

pub struct Arch;

🖥️ 🛠️ Arch-related functionality.

If you are on this architecture, here is what the CPU itself can do.

Methods

• Architecture-specific instructions
  • Portable abstractions
  • x86 or x86_64 instructions
  • arm instructions
  • AArch64 instructions
  • riscv instructions
• Implementations that depend on: dep_safe_arch, (x86 or x86_64) and some target feature:
  • none.
  • adx.
  • aes.
  • avx.
  • avx2.
  • bmi1.
  • bmi2.
  • fma.
  • lzcnt.
  • pclmulqdq.
  • popcnt.
  • rdrand.
  • rdseed.
  • sse (generic).
  • sse2.
  • sse3.
  • sse4.1.
  • sse4.2.
  • ssse3.

Implementations

impl Arch

Portable abstractions over architecture-dependent instructions.

pub fn cycles() -> u64

Available on crate feature unsafe_hint and (AArch64 or ARM or RISC-V RV32 or RISC-V RV64 or x86 or x86-64) only.

A portable, best-effort raw cycle counter for performance measurement.

Notes and Warnings

• The behavior and availability is entirely architecture-dependent.
• On x86, this uses the TSC. Ensure an ‘invariant TSC’ exists.
• On ARM (32-bit and 64-bit), this uses the Virtual Count Register (CNTVCT).
• On RISC-V, this uses the rdcycle instruction.
• The value is only meaningful for measuring relative durations on the same core.

impl Arch

X86 or X86_64 instructions

pub fn rdtsc() -> u64

Available on crate feature unsafe_hint and (x86 or x86-64) only.

Reads the current value of the processor’s time-stamp counter.

See ::core::arch::x86_64::_rdtsc().

pub fn rdtscp() -> (u64, u32)

Available on crate feature unsafe_hint and (x86 or x86-64) only.

Reads the current value of the processor’s time-stamp counter and the processor ID.

The processor ID:

• is a CPU core identifier (logical/core number), not a process ID.
• is tipically a small u32 (e.g., 0 to N-1 for N cores). See ::core::arch::x86_64::__rdtscp().

impl Arch

ARM instructions

pub fn cntvct() -> u64

Available on ARM and crate feature unsafe_hint only.

Reads the Virtual Count Register (CNTVCT) using the appropriate instruction.

This provides a 64-bit cycle counter that is virtualized and intended for timing purposes. It is the ARM equivalent of x86’s TSC.

Availability

Requires ARM Virtualization Extensions. Common on Cortex-A7, A15, A53, A57, and later.

On modern Linux systems, this register is typically available from user space.

impl Arch

AArch64 instructions

pub fn cntvct() -> u64

Available on AArch64 and crate feature unsafe_hint only.

Reads the Virtual Count Register (CNTVCT_EL0) using mrs cntvct_el0.

This is the standard userspace-accessible cycle counter on AArch64.

impl Arch

This impl block contains no public items.

RISC-V instructions

impl Arch

Functions not requiring any target feature.

---

pub fn byte_swap_i32(i: i32) -> i32

Available on (crate features dep_safe_arch) and (x86 or x86-64) only.

Swap the bytes of the given 32-bit value.

See in docs.rs: byte_swap_i32.

pub fn byte_swap_i64(i: i64) -> i64

Available on (crate features dep_safe_arch) and (x86 or x86-64) only.

Swap the bytes of the given 64-bit value.

See in docs.rs: byte_swap_i64.

pub fn read_timestamp_counter() -> u64

Available on (crate features dep_safe_arch) and (x86 or x86-64) only.

Reads the CPU’s timestamp counter value.

See in docs.rs: read_timestamp_counter.

pub fn read_timestamp_counter_p(aux: &mut u32) -> u64

Available on (crate features dep_safe_arch) and (x86 or x86-64) only.

Reads the CPU’s timestamp counter value and store the processor signature.

See in docs.rs: read_timestamp_counter_p.

impl Arch

Functions requiring the adx target feature.

---

See: https://en.wikipedia.org/wiki/Intel_ADX

pub fn add_carry_u32(c_in: u8, a: u32, b: u32, out: &mut u32) -> u8

Available on (crate features dep_safe_arch) and (target features adx) and (x86 or x86-64) only.

Add two u32 with a carry value.

See in docs.rs: add_carry_u32.

pub fn add_carry_u64(c_in: u8, a: u64, b: u64, out: &mut u64) -> u8

Available on (crate features dep_safe_arch) and (target features adx) and (x86 or x86-64) only.

Add two u64 with a carry value.

See in docs.rs: add_carry_u64.

impl Arch

Functions requiring the aes target feature.

---

See: https://en.wikipedia.org/wiki/AES_instruction_set

pub fn aes_decrypt_last_m128i(a: m128i, round_key: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features aes) and (x86 or x86-64) only.

Perform the last round of an AES decryption flow on a using the round_key.

See in docs.rs: aes_decrypt_last_m128i.

pub fn aes_decrypt_m128i(a: m128i, round_key: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features aes) and (x86 or x86-64) only.

Perform one round of an AES decryption flow on a using the round_key.

See in docs.rs: aes_decrypt_m128i.

pub fn aes_encrypt_last_m128i(a: m128i, round_key: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features aes) and (x86 or x86-64) only.

Perform the last round of an AES encryption flow on a using the round_key.

See in docs.rs: aes_encrypt_last_m128i.

pub fn aes_encrypt_m128i(a: m128i, round_key: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features aes) and (x86 or x86-64) only.

Perform one round of an AES encryption flow on a using the round_key.

See in docs.rs: aes_encrypt_m128i.

pub fn aes_inv_mix_columns_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features aes) and (x86 or x86-64) only.

Perform the InvMixColumns transform on a.

See in docs.rs: aes_inv_mix_columns_m128i.

pub fn aes_key_gen_assist_m128i<const IMM: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features aes) and (x86 or x86-64) only.

Assist in expanding an AES cipher key.

See in docs.rs: aes_key_gen_assist_m128i.

impl Arch

Functions requiring the avx target feature.

---

See: https://en.wikipedia.org/wiki/Advanced_Vector_Extensions

pub fn add_horizontal_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Add adjacent f32 lanes.

See in docs.rs: add_horizontal_m256.

pub fn add_horizontal_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Add adjacent f64 lanes.

See in docs.rs: add_horizontal_m256d.

pub fn add_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise a + b with f32 lanes.

See in docs.rs: add_m256.

pub fn add_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise a + b with f64 lanes.

See in docs.rs: add_m256d.

pub fn addsub_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Alternately, from the top, add f32 then sub f32.

See in docs.rs: addsub_m256.

pub fn addsub_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Alternately, from the top, add f64 then sub f64.

See in docs.rs: addsub_m256d.

pub fn bitand_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bitwise a & b.

See in docs.rs: bitand_m256.

pub fn bitand_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bitwise a & b.

See in docs.rs: bitand_m256d.

pub fn bitandnot_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bitwise (!a) & b.

See in docs.rs: bitandnot_m256.

pub fn bitandnot_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bitwise (!a) & b.

See in docs.rs: bitandnot_m256d.

pub fn bitor_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bitwise a | b.

See in docs.rs: bitor_m256.

pub fn bitor_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bitwise a | b.

See in docs.rs: bitor_m256d.

pub fn bitxor_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bitwise a ^ b.

See in docs.rs: bitxor_m256.

pub fn bitxor_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bitwise a ^ b.

See in docs.rs: bitxor_m256d.

pub fn blend_m256<const IMM: i32>(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Blends the f32 lanes according to the immediate mask.

See in docs.rs: blend_m256.

pub fn blend_m256d<const IMM: i32>(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Blends the f64 lanes according to the immediate mask.

See in docs.rs: blend_m256d.

pub fn blend_varying_m256(a: m256, b: m256, mask: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Blend the lanes according to a runtime varying mask.

See in docs.rs: blend_varying_m256.

pub fn blend_varying_m256d(a: m256d, b: m256d, mask: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Blend the lanes according to a runtime varying mask.

See in docs.rs: blend_varying_m256d.

pub fn cast_to_m128_from_m256(a: m256) -> m128

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bit-preserving cast to m128 from m256.

See in docs.rs: cast_to_m128_from_m256.

pub fn cast_to_m128d_from_m256d(a: m256d) -> m128d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bit-preserving cast to m128d from m256d.

See in docs.rs: cast_to_m128d_from_m256d.

pub fn cast_to_m128i_from_m256i(a: m256i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bit-preserving cast to m128i from m256i.

See in docs.rs: cast_to_m128i_from_m256i.

pub fn cast_to_m256_from_m256d(a: m256d) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bit-preserving cast to m256 from m256d.

See in docs.rs: cast_to_m256_from_m256d.

pub fn cast_to_m256_from_m256i(a: m256i) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bit-preserving cast to m256 from m256i.

See in docs.rs: cast_to_m256_from_m256i.

pub fn cast_to_m256d_from_m256(a: m256) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bit-preserving cast to m256i from m256.

See in docs.rs: cast_to_m256d_from_m256.

pub fn cast_to_m256d_from_m256i(a: m256i) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bit-preserving cast to m256d from m256i.

See in docs.rs: cast_to_m256d_from_m256i.

pub fn cast_to_m256i_from_m256(a: m256) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bit-preserving cast to m256i from m256.

See in docs.rs: cast_to_m256i_from_m256.

pub fn cast_to_m256i_from_m256d(a: m256d) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Bit-preserving cast to m256i from m256d.

See in docs.rs: cast_to_m256i_from_m256d.

pub fn ceil_m256(a: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Round f32 lanes towards positive infinity.

See in docs.rs: ceil_m256.

pub fn ceil_m256d(a: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Round f64 lanes towards positive infinity.

See in docs.rs: ceil_m256d.

pub fn cmp_op_mask_m128<const OP: i32>(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Compare f32 lanes according to the operation specified, mask output.

See in docs.rs: cmp_op_mask_m128.

pub fn cmp_op_mask_m128_s<const OP: i32>(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Compare f32 lanes according to the operation specified, mask output.

See in docs.rs: cmp_op_mask_m128_s.

pub fn cmp_op_mask_m128d<const OP: i32>(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Compare f64 lanes according to the operation specified, mask output.

See in docs.rs: cmp_op_mask_m128d.

pub fn cmp_op_mask_m128d_s<const OP: i32>(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Compare f64 lanes according to the operation specified, mask output.

See in docs.rs: cmp_op_mask_m128d_s.

pub fn cmp_op_mask_m256<const OP: i32>(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Compare f32 lanes according to the operation specified, mask output.

See in docs.rs: cmp_op_mask_m256.

pub fn cmp_op_mask_m256d<const OP: i32>(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Compare f64 lanes according to the operation specified, mask output.

See in docs.rs: cmp_op_mask_m256d.

pub fn convert_to_f32_from_m256_s(a: m256) -> f32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Convert the lowest f32 lane to a single f32.

See in docs.rs: convert_to_f32_from_m256_s.

pub fn convert_to_f64_from_m256d_s(a: m256d) -> f64

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Convert the lowest f64 lane to a single f64.

See in docs.rs: convert_to_f64_from_m256d_s.

pub fn convert_to_i32_from_m256i_s(a: m256i) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Convert the lowest i32 lane to a single i32.

See in docs.rs: convert_to_i32_from_m256i_s.

pub fn convert_to_i32_m128i_from_m256d(a: m256d) -> m128i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Convert f64 lanes to be i32 lanes.

See in docs.rs: convert_to_i32_m128i_from_m256d.

pub fn convert_to_i32_m256i_from_m256(a: m256) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Convert f32 lanes to be i32 lanes.

See in docs.rs: convert_to_i32_m256i_from_m256.

pub fn convert_to_m128_from_m256d(a: m256d) -> m128

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Convert f64 lanes to be f32 lanes.

See in docs.rs: convert_to_m128_from_m256d.

pub fn convert_to_m256_from_i32_m256i(a: m256i) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Convert i32 lanes to be f32 lanes.

See in docs.rs: convert_to_m256_from_i32_m256i.

pub fn convert_to_m256d_from_i32_m128i(a: m128i) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Convert i32 lanes to be f64 lanes.

See in docs.rs: convert_to_m256d_from_i32_m128i.

pub fn convert_to_m256d_from_m128(a: m128) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Convert f32 lanes to be f64 lanes.

See in docs.rs: convert_to_m256d_from_m128.

pub fn convert_truncate_to_i32_m128i_from_m256d(a: m256d) -> m128i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Convert f64 lanes to i32 lanes with truncation.

See in docs.rs: convert_truncate_to_i32_m128i_from_m256d.

pub fn convert_truncate_to_i32_m256i_from_m256(a: m256) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Convert f32 lanes to i32 lanes with truncation.

See in docs.rs: convert_truncate_to_i32_m256i_from_m256.

pub fn div_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise a / b with f32.

See in docs.rs: div_m256.

pub fn div_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise a / b with f64.

See in docs.rs: div_m256d.

pub fn dot_product_m256<const IMM: i32>(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

This works like dot_product_m128, but twice as wide.

See in docs.rs: dot_product_m256.

pub fn duplicate_even_lanes_m256(a: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Duplicate the even-indexed lanes to the odd lanes.

See in docs.rs: duplicate_even_lanes_m256.

pub fn duplicate_odd_lanes_m256(a: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Duplicate the odd-indexed lanes to the even lanes.

See in docs.rs: duplicate_odd_lanes_m256.

pub fn duplicate_odd_lanes_m256d(a: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Duplicate the odd-indexed lanes to the even lanes.

See in docs.rs: duplicate_odd_lanes_m256d.

pub fn extract_i32_from_m256i<const IMM: i32>(a: m256i) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Extracts an i32 lane from m256i

See in docs.rs: extract_i32_from_m256i.

pub fn extract_i64_from_m256i<const IMM: i32>(a: m256i) -> i64

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Extracts an i64 lane from m256i

See in docs.rs: extract_i64_from_m256i.

pub fn extract_m128_from_m256<const IMM: i32>(a: m256) -> m128

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Extracts an m128 from m256

See in docs.rs: extract_m128_from_m256.

pub fn extract_m128d_from_m256d<const IMM: i32>(a: m256d) -> m128d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Extracts an m128d from m256d

See in docs.rs: extract_m128d_from_m256d.

pub fn extract_m128i_from_m256i<const IMM: i32>(a: m256i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Extracts an m128i from m256i

See in docs.rs: extract_m128i_from_m256i.

pub fn floor_m256(a: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Round f32 lanes towards negative infinity.

See in docs.rs: floor_m256.

pub fn floor_m256d(a: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Round f64 lanes towards negative infinity.

See in docs.rs: floor_m256d.

pub fn insert_i16_to_m256i<const IMM: i32>(a: m256i, i: i16) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Inserts an i16 to m256i

See in docs.rs: insert_i16_to_m256i.

pub fn insert_i32_to_m256i<const IMM: i32>(a: m256i, i: i32) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Inserts an i32 to m256i

See in docs.rs: insert_i32_to_m256i.

pub fn insert_i64_to_m256i<const IMM: i32>(a: m256i, i: i64) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Inserts an i64 to m256i

See in docs.rs: insert_i64_to_m256i.

pub fn insert_i8_to_m256i<const IMM: i32>(a: m256i, i: i8) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Inserts an i8 to m256i

See in docs.rs: insert_i8_to_m256i.

pub fn insert_m128_to_m256<const IMM: i32>(a: m256, b: m128) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Inserts an m128 to m256

See in docs.rs: insert_m128_to_m256.

pub fn insert_m128d_to_m256d<const IMM: i32>(a: m256d, b: m128d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Inserts an m128d to m256d

See in docs.rs: insert_m128d_to_m256d.

pub fn insert_m128i_to_m256i_slow_avx<const IMM: i32>( a: m256i, b: m128i, ) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Slowly inserts an m128i to m256i.

See in docs.rs: insert_m128i_to_m256i_slow_avx.

pub fn load_f32_splat_m256(a: &f32) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load an f32 and splat it to all lanes of an m256d

See in docs.rs: load_f32_splat_m256.

pub fn load_f64_splat_m256d(a: &f64) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load an f64 and splat it to all lanes of an m256d

See in docs.rs: load_f64_splat_m256d.

pub fn load_m128_splat_m256(a: &m128) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load an m128 and splat it to the lower and upper half of an m256

See in docs.rs: load_m128_splat_m256.

pub fn load_m128d_splat_m256d(a: &m128d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load an m128d and splat it to the lower and upper half of an m256d

See in docs.rs: load_m128d_splat_m256d.

pub fn load_m256(a: &m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register.

See in docs.rs: load_m256.

pub fn load_m256d(a: &m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register.

See in docs.rs: load_m256d.

pub fn load_m256i(a: &m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register.

See in docs.rs: load_m256i.

pub fn load_masked_m128(a: &m128, mask: m128i) -> m128

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register according to a mask.

See in docs.rs: load_masked_m128.

pub fn load_masked_m128d(a: &m128d, mask: m128i) -> m128d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register according to a mask.

See in docs.rs: load_masked_m128d.

pub fn load_masked_m256(a: &m256, mask: m256i) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register according to a mask.

See in docs.rs: load_masked_m256.

pub fn load_masked_m256d(a: &m256d, mask: m256i) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register according to a mask.

See in docs.rs: load_masked_m256d.

pub fn load_unaligned_hi_lo_m256(a: &[f32; 4], b: &[f32; 4]) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register.

See in docs.rs: load_unaligned_hi_lo_m256.

pub fn load_unaligned_hi_lo_m256d(a: &[f64; 2], b: &[f64; 2]) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register.

See in docs.rs: load_unaligned_hi_lo_m256d.

pub fn load_unaligned_hi_lo_m256i(a: &[i8; 16], b: &[i8; 16]) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register.

See in docs.rs: load_unaligned_hi_lo_m256i.

pub fn load_unaligned_m256(a: &[f32; 8]) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register.

See in docs.rs: load_unaligned_m256.

pub fn load_unaligned_m256d(a: &[f64; 4]) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register.

See in docs.rs: load_unaligned_m256d.

pub fn load_unaligned_m256i(a: &[i8; 32]) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Load data from memory into a register.

See in docs.rs: load_unaligned_m256i.

pub fn max_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise max(a, b).

See in docs.rs: max_m256.

pub fn max_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise max(a, b).

See in docs.rs: max_m256d.

pub fn min_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise min(a, b).

See in docs.rs: min_m256.

pub fn min_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise min(a, b).

See in docs.rs: min_m256d.

pub fn move_mask_m256(a: m256) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Collects the sign bit of each lane into a 4-bit value.

See in docs.rs: move_mask_m256.

pub fn move_mask_m256d(a: m256d) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Collects the sign bit of each lane into a 4-bit value.

See in docs.rs: move_mask_m256d.

pub fn mul_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise a * b with f32 lanes.

See in docs.rs: mul_m256.

pub fn mul_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise a * b with f64 lanes.

See in docs.rs: mul_m256d.

pub fn permute2z_m256<const MASK: i32>(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Shuffle 128 bits of floating point data at a time from $a and $b using an immediate control value.

See in docs.rs: permute2z_m256.

pub fn permute2z_m256d<const MASK: i32>(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Shuffle 128 bits of floating point data at a time from a and b using an immediate control value.

See in docs.rs: permute2z_m256d.

pub fn permute2z_m256i<const MASK: i32>(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Slowly swizzle 128 bits of integer data from a and b using an immediate control value.

See in docs.rs: permute2z_m256i.

pub fn permute_m128<const MASK: i32>(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Shuffle the f32 lanes from a using an immediate control value.

See in docs.rs: permute_m128.

pub fn permute_m128d<const MASK: i32>(a: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Shuffle the f64 lanes in a using an immediate control value.

See in docs.rs: permute_m128d.

pub fn permute_m256<const MASK: i32>(a: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Shuffle the f32 lanes in a using an immediate control value.

See in docs.rs: permute_m256.

pub fn permute_m256d<const MASK: i32>(a: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Shuffle the f64 lanes from a together using an immediate control value.

See in docs.rs: permute_m256d.

pub fn reciprocal_m256(a: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Reciprocal of f32 lanes.

See in docs.rs: reciprocal_m256.

pub fn reciprocal_sqrt_m256(a: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Reciprocal of f32 lanes.

See in docs.rs: reciprocal_sqrt_m256.

pub fn round_m256<const OP: i32>(a: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Rounds each lane in the style specified.

See in docs.rs: round_m256.

pub fn round_m256d<const OP: i32>(a: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Rounds each lane in the style specified.

See in docs.rs: round_m256d.

pub fn set_i16_m256i( e15: i16, e14: i16, e13: i16, e12: i16, e11: i16, e10: i16, e9: i16, e8: i16, e7: i16, e6: i16, e5: i16, e4: i16, e3: i16, e2: i16, e1: i16, e0: i16, ) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set i16 args into an m256i lane.

See in docs.rs: set_i16_m256i.

pub fn set_i32_m256i( e7: i32, e6: i32, e5: i32, e4: i32, e3: i32, e2: i32, e1: i32, e0: i32, ) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set i32 args into an m256i lane.

See in docs.rs: set_i32_m256i.

pub fn set_i64_m256i(e3: i64, e2: i64, e1: i64, e0: i64) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set i64 args into an m256i lane.

See in docs.rs: set_i64_m256i.

pub fn set_i8_m256i( e31: i8, e30: i8, e29: i8, e28: i8, e27: i8, e26: i8, e25: i8, e24: i8, e23: i8, e22: i8, e21: i8, e20: i8, e19: i8, e18: i8, e17: i8, e16: i8, e15: i8, e14: i8, e13: i8, e12: i8, e11: i8, e10: i8, e9: i8, e8: i8, e7: i8, e6: i8, e5: i8, e4: i8, e3: i8, e2: i8, e1: i8, e0: i8, ) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set i8 args into an m256i lane.

See in docs.rs: set_i8_m256i.

pub fn set_m128_m256(high: m128, low: m128) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set m128 args into an m256.

See in docs.rs: set_m128_m256.

pub fn set_m128d_m256d(high: m128d, low: m128d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set m128d args into an m256d.

See in docs.rs: set_m128d_m256d.

pub fn set_m128i_m256i(hi: m128i, lo: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set m128i args into an m256i.

See in docs.rs: set_m128i_m256i.

pub fn set_m256( e7: f32, e6: f32, e5: f32, e4: f32, e3: f32, e2: f32, e1: f32, e0: f32, ) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set f32 args into an m256 lane.

See in docs.rs: set_m256.

pub fn set_m256d(e3: f64, e2: f64, e1: f64, e0: f64) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set f64 args into an m256d lane.

See in docs.rs: set_m256d.

pub fn set_reversed_i16_m256i( e15: i16, e14: i16, e13: i16, e12: i16, e11: i16, e10: i16, e9: i16, e8: i16, e7: i16, e6: i16, e5: i16, e4: i16, e3: i16, e2: i16, e1: i16, e0: i16, ) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set i16 args into an m256i lane.

See in docs.rs: set_reversed_i16_m256i.

pub fn set_reversed_i32_m256i( e7: i32, e6: i32, e5: i32, e4: i32, e3: i32, e2: i32, e1: i32, e0: i32, ) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set i32 args into an m256i lane.

See in docs.rs: set_reversed_i32_m256i.

pub fn set_reversed_i64_m256i(e3: i64, e2: i64, e1: i64, e0: i64) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set i64 args into an m256i lane.

See in docs.rs: set_reversed_i64_m256i.

pub fn set_reversed_i8_m256i( e31: i8, e30: i8, e29: i8, e28: i8, e27: i8, e26: i8, e25: i8, e24: i8, e23: i8, e22: i8, e21: i8, e20: i8, e19: i8, e18: i8, e17: i8, e16: i8, e15: i8, e14: i8, e13: i8, e12: i8, e11: i8, e10: i8, e9: i8, e8: i8, e7: i8, e6: i8, e5: i8, e4: i8, e3: i8, e2: i8, e1: i8, e0: i8, ) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set i8 args into an m256i lane.

See in docs.rs: set_reversed_i8_m256i.

pub fn set_reversed_m128_m256(hi: m128, lo: m128) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set m128 args into an m256.

See in docs.rs: set_reversed_m128_m256.

pub fn set_reversed_m128d_m256d(hi: m128d, lo: m128d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set m128d args into an m256d.

See in docs.rs: set_reversed_m128d_m256d.

pub fn set_reversed_m128i_m256i(hi: m128i, lo: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set m128i args into an m256i.

See in docs.rs: set_reversed_m128i_m256i.

pub fn set_reversed_m256( e7: f32, e6: f32, e5: f32, e4: f32, e3: f32, e2: f32, e1: f32, e0: f32, ) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set f32 args into an m256 lane.

See in docs.rs: set_reversed_m256.

pub fn set_reversed_m256d(e3: f64, e2: f64, e1: f64, e0: f64) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Set f64 args into an m256d lane.

See in docs.rs: set_reversed_m256d.

pub fn set_splat_i16_m256i(i: i16) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Splat an i16 arg into an m256i lane.

See in docs.rs: set_splat_i16_m256i.

pub fn set_splat_i32_m256i(i: i32) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Splat an i32 arg into an m256i lane.

See in docs.rs: set_splat_i32_m256i.

pub fn set_splat_i64_m256i(i: i64) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Splat an i64 arg into an m256i lane.

See in docs.rs: set_splat_i64_m256i.

pub fn set_splat_i8_m256i(i: i8) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Splat an i8 arg into an m256i lane.

See in docs.rs: set_splat_i8_m256i.

pub fn set_splat_m256(f: f32) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Splat an f32 arg into an m256 lane.

See in docs.rs: set_splat_m256.

pub fn set_splat_m256d(f: f64) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Splat an f64 arg into an m256d lane.

See in docs.rs: set_splat_m256d.

pub fn shuffle_av_f32_all_m128(a: m128, v: m128i) -> m128

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Shuffle f32 values in a using i32 values in v.

See in docs.rs: shuffle_av_f32_all_m128.

pub fn shuffle_av_f32_half_m256(a: m256, v: m256i) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Shuffle f32 values in a using i32 values in v.

See in docs.rs: shuffle_av_f32_half_m256.

pub fn shuffle_av_f64_all_m128d(a: m128d, v: m128i) -> m128d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Shuffle f64 lanes in a using bit 1 of the i64 lanes in v

See in docs.rs: shuffle_av_f64_all_m128d.

pub fn shuffle_av_f64_half_m256d(a: m256d, b: m256i) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Shuffle f64 lanes in a using bit 1 of the i64 lanes in v.

See in docs.rs: shuffle_av_f64_half_m256d.

pub fn shuffle_m256<const IMM: i32>(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Shuffle the f32 lanes from a and b together using an immediate control value.

See in docs.rs: shuffle_m256.

pub fn shuffle_m256d<const IMM: i32>(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Shuffle the f64 lanes from a and b together using an immediate control value.

See in docs.rs: shuffle_m256d.

pub fn sqrt_m256(a: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise sqrt on f64 lanes.

See in docs.rs: sqrt_m256.

pub fn sqrt_m256d(a: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise sqrt on f64 lanes.

See in docs.rs: sqrt_m256d.

pub fn store_m256(addr: &mut m256, a: m256)

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory.

See in docs.rs:store_m256.

pub fn store_m256d(addr: &mut m256d, a: m256d)

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory.

See in docs.rs:store_m256d.

pub fn store_m256i(addr: &mut m256i, a: m256i)

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory.

See in docs.rs:store_m256i.

pub fn store_masked_m128(addr: &mut m128, mask: m128i, a: m128)

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory according to a mask.

See in docs.rs:store_masked_m128.

pub fn store_masked_m128d(addr: &mut m128d, mask: m128i, a: m128d)

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory according to a mask.

See in docs.rs:store_masked_m128d.

pub fn store_masked_m256(addr: &mut m256, mask: m256i, a: m256)

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory according to a mask.

See in docs.rs:store_masked_m256.

pub fn store_masked_m256d(addr: &mut m256d, mask: m256i, a: m256d)

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory according to a mask.

See in docs.rs:store_masked_m256d.

pub fn store_unaligned_hi_lo_m256( hi_addr: &mut [f32; 4], lo_addr: &mut [f32; 4], a: m256, )

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory.

See in docs.rs:store_unaligned_hi_lo_m256.

pub fn store_unaligned_hi_lo_m256d( hi_addr: &mut [f64; 2], lo_addr: &mut [f64; 2], a: m256d, )

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory.

See in docs.rs:store_unaligned_hi_lo_m256d.

pub fn store_unaligned_hi_lo_m256i( hi_addr: &mut [i8; 16], lo_addr: &mut [i8; 16], a: m256i, )

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory.

See in docs.rs:store_unaligned_hi_lo_m256i.

pub fn store_unaligned_m256(addr: &mut [f32; 8], a: m256)

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory.

See in docs.rs:store_unaligned_m256.

pub fn store_unaligned_m256d(addr: &mut [f64; 4], a: m256d)

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory.

See in docs.rs:store_unaligned_m256d.

pub fn store_unaligned_m256i(addr: &mut [i8; 32], a: m256i)

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Store data from a register into memory.

See in docs.rs:store_unaligned_m256i.

pub fn sub_horizontal_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Subtract adjacent f32 lanes.

See in docs.rs: sub_horizontal_m256.

pub fn sub_horizontal_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Subtract adjacent f64 lanes.

See in docs.rs: sub_horizontal_m256d.

pub fn sub_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise a - b with f32 lanes.

See in docs.rs: sub_m256.

pub fn sub_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Lanewise a - b with f64 lanes.

See in docs.rs: sub_m256d.

pub fn testc_m128(a: m128, b: m128) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Compute the bitwise of sign bit NOT of a and then AND with b, returns 1 if the result is zero, otherwise 0.

See in docs.rs: testc_m128.

pub fn testc_m128d(a: m128d, b: m128d) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Compute the bitwise of sign bit NOT of a and then AND with b, returns 1 if the result is zero, otherwise 0.

See in docs.rs: testc_m128d.

pub fn testc_m256(a: m256, b: m256) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Compute the bitwise of sign bit NOT of a and then AND with b, returns 1 if the result is zero, otherwise 0.

See in docs.rs: testc_m256.

pub fn testc_m256d(a: m256d, b: m256d) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Compute the bitwise of sign bit NOT of a and then AND with b, returns 1 if the result is zero, otherwise 0.

See in docs.rs: testc_m256d.

pub fn testc_m256i(a: m256i, b: m256i) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Compute the bitwise NOT of a and then AND with b, returns 1 if the result is zero, otherwise 0.

See in docs.rs: testc_m256i.

pub fn testz_m128(a: m128, b: m128) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Computes the bitwise AND of 256 bits in a and b, returns 1 if the result is zero, otherwise 0.

See in docs.rs: testz_m128.

pub fn testz_m128d(a: m128d, b: m128d) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Computes the bitwise of sign bitAND of 256 bits in a and b, returns 1 if the result is zero, otherwise 0.

See in docs.rs: testz_m128d.

pub fn testz_m256(a: m256, b: m256) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Computes the bitwise AND of 256 bits in a and b, returns 1 if the result is zero, otherwise 0.

See in docs.rs: testz_m256.

pub fn testz_m256d(a: m256d, b: m256d) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Computes the bitwise of sign bit AND of 256 bits in a and b, returns 1 if the result is zero, otherwise 0.

See in docs.rs: testz_m256d.

pub fn testz_m256i(a: m256i, b: m256i) -> i32

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Computes the bitwise of sign bit AND of 256 bits in a and b, returns 1 if the result is zero, otherwise 0.

See in docs.rs: testz_m256i.

pub fn unpack_hi_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Unpack and interleave the high lanes.

See in docs.rs: unpack_hi_m256.

pub fn unpack_hi_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Unpack and interleave the high lanes.

See in docs.rs: unpack_hi_m256d.

pub fn unpack_lo_m256(a: m256, b: m256) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Unpack and interleave the high lanes.

See in docs.rs: unpack_lo_m256.

pub fn unpack_lo_m256d(a: m256d, b: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Unpack and interleave the high lanes.

See in docs.rs: unpack_lo_m256d.

pub fn zero_extend_m128(a: m128) -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Zero extend an m128 to m256

See in docs.rs: zero_extend_m128.

pub fn zero_extend_m128d(a: m128d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Zero extend an m128d to m256d

See in docs.rs: zero_extend_m128d.

pub fn zero_extend_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

Zero extend an m128i to m256i

See in docs.rs: zero_extend_m128i.

pub fn zeroed_m256() -> m256

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

A zeroed m256

See in docs.rs: zeroed_m256.

pub fn zeroed_m256d() -> m256d

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

A zeroed m256d

See in docs.rs: zeroed_m256d.

pub fn zeroed_m256i() -> m256i

Available on (crate features dep_safe_arch) and (target features avx) and (x86 or x86-64) only.

A zeroed m256i

See in docs.rs: zeroed_m256i.

impl Arch

Functions requiring the avx2 target feature.

---

See: https://en.wikipedia.org/wiki/Advanced_Vector_Extensions

pub fn abs_i16_m256i(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Absolute value of i16 lanes.

See in docs.rs: abs_i16_m256i.

pub fn abs_i32_m256i(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Absolute value of i32 lanes.

See in docs.rs: abs_i32_m256i.

pub fn abs_i8_m256i(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Absolute value of i8 lanes.

See in docs.rs: abs_i8_m256i.

pub fn add_horizontal_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Horizontal a + b with lanes as i16.

See in docs.rs: add_horizontal_i16_m256i.

pub fn add_horizontal_i32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Horizontal a + b with lanes as i32.

See in docs.rs: add_horizontal_i32_m256i.

pub fn add_horizontal_saturating_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Horizontal saturating a + b with lanes as i16.

See in docs.rs: add_horizontal_saturating_i16_m256i.

pub fn add_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise a + b with lanes as i16.

See in docs.rs: add_i16_m256i.

pub fn add_i32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise a + b with lanes as i32.

See in docs.rs: add_i32_m256i.

pub fn add_i64_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise a + b with lanes as i64.

See in docs.rs: add_i64_m256i.

pub fn add_i8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise a + b with lanes as i8.

See in docs.rs: add_i8_m256i.

pub fn add_saturating_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise saturating a + b with lanes as i16.

See in docs.rs: add_saturating_i16_m256i.

pub fn add_saturating_i8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise saturating a + b with lanes as i8.

See in docs.rs: add_saturating_i8_m256i.

pub fn add_saturating_u16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise saturating a + b with lanes as u16.

See in docs.rs: add_saturating_u16_m256i.

pub fn add_saturating_u8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise saturating a + b with lanes as u8.

See in docs.rs: add_saturating_u8_m256i.

pub fn average_u16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Average u16 lanes.

See in docs.rs: average_u16_m256i.

pub fn average_u8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Average u8 lanes.

See in docs.rs: average_u8_m256i.

pub fn bitand_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Bitwise a & b.

See in docs.rs: bitand_m256i.

pub fn bitandnot_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Bitwise (!a) & b.

See in docs.rs: bitandnot_m256i.

pub fn bitor_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Bitwise a | b

See in docs.rs: bitor_m256i.

pub fn bitxor_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Bitwise a ^ b.

See in docs.rs: bitxor_m256i.

pub fn blend_imm_i16_m256i<const IMM: i32>(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Blends the i16 lanes according to the immediate value.

See in docs.rs: blend_imm_i16_m256i.

pub fn blend_imm_i32_m128i<const IMM: i32>(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Blends the i32 lanes in a and b into a single value.

See in docs.rs: blend_imm_i32_m128i.

pub fn blend_imm_i32_m256i<const IMM: i32>(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Blends the i32 lanes according to the immediate value.

See in docs.rs: blend_imm_i32_m256i.

pub fn blend_varying_i8_m256i(a: m256i, b: m256i, mask: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Blend i8 lanes according to a runtime varying mask.

See in docs.rs: blend_varying_i8_m256i.

pub fn byte_shl_imm_u128_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shifts each u128 lane left by a number of bytes.

See in docs.rs: byte_shl_imm_u128_m256i.

pub fn byte_shr_imm_u128_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shifts each u128 lane right by a number of bytes.

See in docs.rs: byte_shr_imm_u128_m256i.

pub fn cmp_eq_mask_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Compare i16 lanes for equality, mask output.

See in docs.rs: cmp_eq_mask_i16_m256i.

pub fn cmp_eq_mask_i32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Compare i32 lanes for equality, mask output.

See in docs.rs: cmp_eq_mask_i32_m256i.

pub fn cmp_eq_mask_i64_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Compare i64 lanes for equality, mask output.

See in docs.rs: cmp_eq_mask_i64_m256i.

pub fn cmp_eq_mask_i8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Compare i8 lanes for equality, mask output.

See in docs.rs: cmp_eq_mask_i8_m256i.

pub fn cmp_gt_mask_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Compare i16 lanes for a > b, mask output.

See in docs.rs: cmp_gt_mask_i16_m256i.

pub fn cmp_gt_mask_i32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Compare i32 lanes for a > b, mask output.

See in docs.rs: cmp_gt_mask_i32_m256i.

pub fn cmp_gt_mask_i64_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Compare i64 lanes for a > b, mask output.

See in docs.rs: cmp_gt_mask_i64_m256i.

pub fn cmp_gt_mask_i8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Compare i8 lanes for a > b, mask output.

See in docs.rs: cmp_gt_mask_i8_m256i.

pub fn combined_byte_shr_imm_m256i<const IMM: i32>(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Works like combined_byte_shr_imm_m128i, but twice as wide.

See in docs.rs: combined_byte_shr_imm_m256i.

pub fn convert_to_i16_m256i_from_i8_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Convert i8 values to i16 values.

See in docs.rs: convert_to_i16_m256i_from_i8_m128i.

pub fn convert_to_i16_m256i_from_lower4_u8_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Convert lower 4 u8 values to i16 values.

See in docs.rs: convert_to_i16_m256i_from_lower4_u8_m128i.

pub fn convert_to_i16_m256i_from_lower8_u8_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Convert lower 8 u8 values to i16 values.

See in docs.rs: convert_to_i16_m256i_from_lower8_u8_m128i.

pub fn convert_to_i16_m256i_from_u8_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Convert u8 values to i16 values.

See in docs.rs: convert_to_i16_m256i_from_u8_m128i.

pub fn convert_to_i32_m256i_from_i16_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Convert i16 values to i32 values.

See in docs.rs: convert_to_i32_m256i_from_i16_m128i.

pub fn convert_to_i32_m256i_from_lower8_i8_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Convert the lower 8 i8 values to i32 values.

See in docs.rs: convert_to_i32_m256i_from_lower8_i8_m128i.

pub fn convert_to_i32_m256i_from_u16_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Convert u16 values to i32 values.

See in docs.rs: convert_to_i32_m256i_from_u16_m128i.

pub fn convert_to_i64_m256i_from_i32_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Convert i32 values to i64 values.

See in docs.rs: convert_to_i64_m256i_from_i32_m128i.

pub fn convert_to_i64_m256i_from_lower4_i16_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Convert i16 values to i64 values.

See in docs.rs: convert_to_i64_m256i_from_lower4_i16_m128i.

pub fn convert_to_i64_m256i_from_lower4_i8_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Convert the lower 4 i8 values to i64 values.

See in docs.rs: convert_to_i64_m256i_from_lower4_i8_m128i.

pub fn convert_to_i64_m256i_from_lower4_u16_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Convert u16 values to i64 values.

See in docs.rs: convert_to_i64_m256i_from_lower4_u16_m128i.

pub fn convert_to_i64_m256i_from_u32_m128i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Convert u32 values to i64 values.

See in docs.rs: convert_to_i64_m256i_from_u32_m128i.

pub fn extract_i16_as_i32_m256i<const LANE: i32>(a: m256i) -> i32

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Gets an i16 value out of an m256i, returns as i32.

See in docs.rs: extract_i16_as_i32_m256i.

pub fn extract_i8_as_i32_m256i<const LANE: i32>(a: m256i) -> i32

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Gets an i8 value out of an m256i, returns as i32.

See in docs.rs: extract_i8_as_i32_m256i.

pub fn extract_m128i_m256i<const LANE: i32>(a: m256i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Gets an m128i value out of an m256i.

See in docs.rs: extract_m128i_m256i.

pub fn insert_m128i_to_m256i<const LANE: i32>(a: m256i, b: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Inserts an m128i to an m256i at the high or low position.

See in docs.rs: insert_m128i_to_m256i.

pub fn load_masked_i32_m128i(a: &m128i, mask: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Loads the reference given and zeroes any i32 lanes not in the mask.

See in docs.rs: load_masked_i32_m128i.

pub fn load_masked_i32_m256i(a: &m256i, mask: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Loads the reference given and zeroes any i32 lanes not in the mask.

See in docs.rs: load_masked_i32_m256i.

pub fn load_masked_i64_m128i(a: &m128i, mask: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Loads the reference given and zeroes any i64 lanes not in the mask.

See in docs.rs: load_masked_i64_m128i.

pub fn load_masked_i64_m256i(a: &m256i, mask: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Loads the reference given and zeroes any i64 lanes not in the mask.

See in docs.rs: load_masked_i64_m256i.

pub fn max_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise max(a, b) with lanes as i16.

See in docs.rs: max_i16_m256i.

pub fn max_i32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise max(a, b) with lanes as i32.

See in docs.rs: max_i32_m256i.

pub fn max_i8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise max(a, b) with lanes as i8.

See in docs.rs: max_i8_m256i.

pub fn max_u16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise max(a, b) with lanes as u16.

See in docs.rs: max_u16_m256i.

pub fn max_u32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise max(a, b) with lanes as u32.

See in docs.rs: max_u32_m256i.

pub fn max_u8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise max(a, b) with lanes as u8.

See in docs.rs: max_u8_m256i.

pub fn min_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise min(a, b) with lanes as i16.

See in docs.rs: min_i16_m256i.

pub fn min_i32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise min(a, b) with lanes as i32.

See in docs.rs: min_i32_m256i.

pub fn min_i8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise min(a, b) with lanes as i8.

See in docs.rs: min_i8_m256i.

pub fn min_u16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise min(a, b) with lanes as u16.

See in docs.rs: min_u16_m256i.

pub fn min_u32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise min(a, b) with lanes as u32.

See in docs.rs: min_u32_m256i.

pub fn min_u8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise min(a, b) with lanes as u8.

See in docs.rs: min_u8_m256i.

pub fn move_mask_i8_m256i(a: m256i) -> i32

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Create an i32 mask of each sign bit in the i8 lanes.

See in docs.rs: move_mask_i8_m256i.

pub fn mul_i16_horizontal_add_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Multiply i16 lanes producing i32 values, horizontal add pairs of i32 values to produce the final output.

See in docs.rs: mul_i16_horizontal_add_m256i.

pub fn mul_i16_keep_high_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Multiply the i16 lanes and keep the high half of each 32-bit output.

See in docs.rs: mul_i16_keep_high_m256i.

pub fn mul_i16_keep_low_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Multiply the i16 lanes and keep the low half of each 32-bit output.

See in docs.rs: mul_i16_keep_low_m256i.

pub fn mul_i16_scale_round_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Multiply i16 lanes into i32 intermediates, keep the high 18 bits, round by adding 1, right shift by 1.

See in docs.rs: mul_i16_scale_round_m256i.

pub fn mul_i32_keep_low_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Multiply the i32 lanes and keep the low half of each 64-bit output.

See in docs.rs: mul_i32_keep_low_m256i.

pub fn mul_i64_low_bits_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Multiply the lower i32 within each i64 lane, i64 output.

See in docs.rs: mul_i64_low_bits_m256i.

pub fn mul_u16_keep_high_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Multiply the u16 lanes and keep the high half of each 32-bit output.

See in docs.rs: mul_u16_keep_high_m256i.

pub fn mul_u64_low_bits_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Multiply the lower u32 within each u64 lane, u64 output.

See in docs.rs: mul_u64_low_bits_m256i.

pub fn mul_u8i8_add_horizontal_saturating_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

This is dumb and weird.

See in docs.rs: mul_u8i8_add_horizontal_saturating_m256i.

pub fn multi_packed_sum_abs_diff_u8_m256i<const IMM: i32>( a: m256i, b: m256i, ) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Computes eight u16 “sum of absolute difference” values according to the bytes selected.

See in docs.rs: multi_packed_sum_abs_diff_u8_m256i.

pub fn pack_i16_to_i8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Saturating convert i16 to i8, and pack the values.

See in docs.rs: pack_i16_to_i8_m256i.

pub fn pack_i16_to_u8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Saturating convert i16 to u8, and pack the values.

See in docs.rs: pack_i16_to_u8_m256i.

pub fn pack_i32_to_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Saturating convert i32 to i16, and pack the values.

See in docs.rs: pack_i32_to_i16_m256i.

pub fn pack_i32_to_u16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Saturating convert i32 to u16, and pack the values.

See in docs.rs: pack_i32_to_u16_m256i.

pub fn set_splat_i16_m128i_s_m256i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Sets the lowest i16 lane of an m128i as all lanes of an m256i.

See in docs.rs: set_splat_i16_m128i_s_m256i.

pub fn set_splat_i32_m128i_s_m256i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Sets the lowest i32 lane of an m128i as all lanes of an m256i.

See in docs.rs: set_splat_i32_m128i_s_m256i.

pub fn set_splat_i64_m128i_s_m256i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Sets the lowest i64 lane of an m128i as all lanes of an m256i.

See in docs.rs: set_splat_i64_m128i_s_m256i.

pub fn set_splat_i8_m128i_s_m256i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Sets the lowest i8 lane of an m128i as all lanes of an m256i.

See in docs.rs: set_splat_i8_m128i_s_m256i.

pub fn set_splat_m128_s_m256(a: m128) -> m256

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Sets the lowest lane of an m128 as all lanes of an m256.

See in docs.rs: set_splat_m128_s_m256.

pub fn set_splat_m128d_s_m256d(a: m128d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Sets the lowest lane of an m128d as all lanes of an m256d.

See in docs.rs: set_splat_m128d_s_m256d.

pub fn shl_all_u16_m256i(a: m256i, count: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise u16 shift left by the lower u64 lane of count.

See in docs.rs: shl_all_u16_m256i.

pub fn shl_all_u32_m256i(a: m256i, count: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shift all u32 lanes left by the lower u64 lane of count.

See in docs.rs: shl_all_u32_m256i.

pub fn shl_all_u64_m256i(a: m256i, count: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shift all u64 lanes left by the lower u64 lane of count.

See in docs.rs: shl_all_u64_m256i.

pub fn shl_each_u32_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shift u32 values to the left by count bits.

See in docs.rs: shl_each_u32_m128i.

pub fn shl_each_u32_m256i(a: m256i, count: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise u32 shift left by the matching i32 lane in count.

See in docs.rs: shl_each_u32_m256i.

pub fn shl_each_u64_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shift u64 values to the left by count bits.

See in docs.rs: shl_each_u64_m128i.

pub fn shl_each_u64_m256i(a: m256i, count: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise u64 shift left by the matching u64 lane in count.

See in docs.rs: shl_each_u64_m256i.

pub fn shl_imm_u16_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shifts all u16 lanes left by an immediate.

See in docs.rs: shl_imm_u16_m256i.

pub fn shl_imm_u32_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shifts all u32 lanes left by an immediate.

See in docs.rs: shl_imm_u32_m256i.

pub fn shl_imm_u64_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shifts all u64 lanes left by an immediate.

See in docs.rs: shl_imm_u64_m256i.

pub fn shr_all_i16_m256i(a: m256i, count: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise i16 shift right by the lower i64 lane of count.

See in docs.rs: shr_all_i16_m256i.

pub fn shr_all_i32_m256i(a: m256i, count: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise i32 shift right by the lower i64 lane of count.

See in docs.rs: shr_all_i32_m256i.

pub fn shr_all_u16_m256i(a: m256i, count: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise u16 shift right by the lower u64 lane of count.

See in docs.rs: shr_all_u16_m256i.

pub fn shr_all_u32_m256i(a: m256i, count: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise u32 shift right by the lower u64 lane of count.

See in docs.rs: shr_all_u32_m256i.

pub fn shr_all_u64_m256i(a: m256i, count: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise u64 shift right by the lower u64 lane of count.

See in docs.rs: shr_all_u64_m256i.

pub fn shr_each_i32_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shift i32 values to the right by count bits.

See in docs.rs: shr_each_i32_m128i.

pub fn shr_each_i32_m256i(a: m256i, count: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise i32 shift right by the matching i32 lane in count.

See in docs.rs: shr_each_i32_m256i.

pub fn shr_each_u32_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shift u32 values to the left by count bits.

See in docs.rs: shr_each_u32_m128i.

pub fn shr_each_u32_m256i(a: m256i, count: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise u32 shift right by the matching u32 lane in count.

See in docs.rs: shr_each_u32_m256i.

pub fn shr_each_u64_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shift u64 values to the left by count bits.

See in docs.rs: shr_each_u64_m128i.

pub fn shr_each_u64_m256i(a: m256i, count: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise u64 shift right by the matching i64 lane in count.

See in docs.rs: shr_each_u64_m256i.

pub fn shr_imm_i16_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shifts all i16 lanes left by an immediate.

See in docs.rs: shr_imm_i16_m256i.

pub fn shr_imm_i32_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shifts all i32 lanes left by an immediate.

See in docs.rs: shr_imm_i32_m256i.

pub fn shr_imm_u16_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shifts all u16 lanes right by an immediate.

See in docs.rs: shr_imm_u16_m256i.

pub fn shr_imm_u32_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shifts all u32 lanes right by an immediate.

See in docs.rs: shr_imm_u32_m256i.

pub fn shr_imm_u64_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shifts all u64 lanes right by an immediate.

See in docs.rs: shr_imm_u64_m256i.

pub fn shuffle_abi_i128z_all_m256i<const MASK: i32>(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shuffle 128 bits of integer data from $a and $b using an immediate control value.

See in docs.rs: shuffle_abi_i128z_all_m256i.

pub fn shuffle_ai_f64_all_m256d<const IMM: i32>(a: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shuffle the f64 lanes from $a using an immediate control value.

See in docs.rs: shuffle_ai_f64_all_m256d.

pub fn shuffle_ai_i16_h64half_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shuffle the high i16 lanes in $a using an immediate control value.

See in docs.rs: shuffle_ai_i16_h64half_m256i.

pub fn shuffle_ai_i16_l64half_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shuffle the low i16 lanes in $a using an immediate control value.

See in docs.rs: shuffle_ai_i16_l64half_m256i.

pub fn shuffle_ai_i32_half_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shuffle the i32 lanes in a using an immediate control value.

See in docs.rs: shuffle_ai_i32_half_m256i.

pub fn shuffle_ai_i64_all_m256i<const IMM: i32>(a: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shuffle the f64 lanes in $a using an immediate control value.

See in docs.rs: shuffle_ai_i64_all_m256i.

pub fn shuffle_av_f32_all_m256(a: m256, v: m256i) -> m256

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shuffle f32 lanes in a using i32 values in v.

See in docs.rs: shuffle_av_f32_all_m256.

pub fn shuffle_av_i32_all_m256i(a: m256i, v: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shuffle i32 lanes in a using i32 values in v.

See in docs.rs: shuffle_av_i32_all_m256i.

pub fn shuffle_av_i8z_half_m256i(a: m256i, v: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Shuffle i8 lanes in a using i8 values in v.

See in docs.rs: shuffle_av_i8z_half_m256i.

pub fn sign_apply_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise a * signum(b) with lanes as i16

See in docs.rs: sign_apply_i16_m256i.

pub fn sign_apply_i32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise a * signum(b) with lanes as i32

See in docs.rs: sign_apply_i32_m256i.

pub fn sign_apply_i8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise a * signum(b) with lanes as i8

See in docs.rs: sign_apply_i8_m256i.

pub fn splat_i16_m128i_s_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Splat the lowest 16-bit lane across the entire 128 bits.

See in docs.rs: splat_i16_m128i_s_m128i.

pub fn splat_i32_m128i_s_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Splat the lowest 32-bit lane across the entire 128 bits.

See in docs.rs: splat_i32_m128i_s_m128i.

pub fn splat_i64_m128i_s_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Splat the lowest 64-bit lane across the entire 128 bits.

See in docs.rs: splat_i64_m128i_s_m128i.

pub fn splat_i8_m128i_s_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Splat the lowest 8-bit lane across the entire 128 bits.

See in docs.rs: splat_i8_m128i_s_m128i.

pub fn splat_m128_s_m128(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Splat the lowest f32 across all four lanes.

See in docs.rs: splat_m128_s_m128.

pub fn splat_m128d_s_m128d(a: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Splat the lower f64 across both lanes of m128d.

See in docs.rs: splat_m128d_s_m128d.

pub fn splat_m128i_m256i(a: m128i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Splat the 128-bits across 256-bits.

See in docs.rs: splat_m128i_m256i.

pub fn store_masked_i32_m128i(addr: &mut m128i, mask: m128i, a: m128i)

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Stores the i32 masked lanes given to the reference.

See in docs.rs:store_masked_i32_m128i.

pub fn store_masked_i32_m256i(addr: &mut m256i, mask: m256i, a: m256i)

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Stores the i32 masked lanes given to the reference.

See in docs.rs:store_masked_i32_m256i.

pub fn store_masked_i64_m128i(addr: &mut m128i, mask: m128i, a: m128i)

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Stores the i32 masked lanes given to the reference.

See in docs.rs:store_masked_i64_m128i.

pub fn store_masked_i64_m256i(addr: &mut m256i, mask: m256i, a: m256i)

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Stores the i32 masked lanes given to the reference.

See in docs.rs:store_masked_i64_m256i.

pub fn sub_horizontal_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Horizontal a - b with lanes as i16.

See in docs.rs: sub_horizontal_i16_m256i.

pub fn sub_horizontal_i32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Horizontal a - b with lanes as i32.

See in docs.rs: sub_horizontal_i32_m256i.

pub fn sub_horizontal_saturating_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Horizontal saturating a - b with lanes as i16.

See in docs.rs: sub_horizontal_saturating_i16_m256i.

pub fn sub_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise a - b with lanes as i16.

See in docs.rs: sub_i16_m256i.

pub fn sub_i32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise a - b with lanes as i32.

See in docs.rs: sub_i32_m256i.

pub fn sub_i64_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise a - b with lanes as i64.

See in docs.rs: sub_i64_m256i.

pub fn sub_i8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise a - b with lanes as i8.

See in docs.rs: sub_i8_m256i.

pub fn sub_saturating_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise saturating a - b with lanes as i16.

See in docs.rs: sub_saturating_i16_m256i.

pub fn sub_saturating_i8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise saturating a - b with lanes as i8.

See in docs.rs: sub_saturating_i8_m256i.

pub fn sub_saturating_u16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise saturating a - b with lanes as u16.

See in docs.rs: sub_saturating_u16_m256i.

pub fn sub_saturating_u8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Lanewise saturating a - b with lanes as u8.

See in docs.rs: sub_saturating_u8_m256i.

pub fn sum_of_u8_abs_diff_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Compute “sum of u8 absolute differences”.

See in docs.rs: sum_of_u8_abs_diff_m256i.

pub fn unpack_high_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Unpack and interleave high i16 lanes of a and b.

See in docs.rs: unpack_high_i16_m256i.

pub fn unpack_high_i32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Unpack and interleave high i32 lanes of a and b.

See in docs.rs: unpack_high_i32_m256i.

pub fn unpack_high_i64_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Unpack and interleave high i64 lanes of a and b.

See in docs.rs: unpack_high_i64_m256i.

pub fn unpack_high_i8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Unpack and interleave high i8 lanes of a and b.

See in docs.rs: unpack_high_i8_m256i.

pub fn unpack_low_i16_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Unpack and interleave low i16 lanes of a and b.

See in docs.rs: unpack_low_i16_m256i.

pub fn unpack_low_i32_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Unpack and interleave low i32 lanes of a and b.

See in docs.rs: unpack_low_i32_m256i.

pub fn unpack_low_i64_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Unpack and interleave low i64 lanes of a and b.

See in docs.rs: unpack_low_i64_m256i.

pub fn unpack_low_i8_m256i(a: m256i, b: m256i) -> m256i

Available on (crate features dep_safe_arch) and (target features avx2) and (x86 or x86-64) only.

Unpack and interleave low i8 lanes of a and b.

See in docs.rs: unpack_low_i8_m256i.

impl Arch

Functions requiring the bmi1 target feature.

---

See: https://en.wikipedia.org/wiki/X86_Bit_manipulation_instruction_set#BMI1_(Bit_Manipulation_Instruction_Set_1)

pub fn bit_extract2_u32(a: u32, control: u32) -> u32

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Extract a span of bits from the u32, control value style.

See in docs.rs: bit_extract2_u32.

pub fn bit_extract2_u64(a: u64, control: u64) -> u64

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Extract a span of bits from the u64, control value style.

See in docs.rs: bit_extract2_u64.

pub fn bit_extract_u32(a: u32, start: u32, len: u32) -> u32

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Extract a span of bits from the u32, start and len style.

See in docs.rs: bit_extract_u32.

pub fn bit_extract_u64(a: u64, start: u32, len: u32) -> u64

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Extract a span of bits from the u64, start and len style.

See in docs.rs: bit_extract_u64.

pub fn bit_lowest_set_mask_u32(a: u32) -> u32

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Gets the mask of all bits up to and including the lowest set bit in a u32.

See in docs.rs: bit_lowest_set_mask_u32.

pub fn bit_lowest_set_mask_u64(a: u64) -> u64

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Gets the mask of all bits up to and including the lowest set bit in a u64.

See in docs.rs: bit_lowest_set_mask_u64.

pub fn bit_lowest_set_reset_u32(a: u32) -> u32

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Resets (clears) the lowest set bit.

See in docs.rs: bit_lowest_set_reset_u32.

pub fn bit_lowest_set_reset_u64(a: u64) -> u64

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Resets (clears) the lowest set bit.

See in docs.rs: bit_lowest_set_reset_u64.

pub fn bit_lowest_set_value_u32(a: u32) -> u32

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Gets the value of the lowest set bit in a u32.

See in docs.rs: bit_lowest_set_value_u32.

pub fn bit_lowest_set_value_u64(a: u64) -> u64

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Gets the value of the lowest set bit in a u64.

See in docs.rs: bit_lowest_set_value_u64.

pub fn bitandnot_u32(a: u32, b: u32) -> u32

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Bitwise (!a) & b for u32

See in docs.rs: bitandnot_u32.

pub fn bitandnot_u64(a: u64, b: u64) -> u64

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Bitwise (!a) & b for u64

See in docs.rs: bitandnot_u64.

pub fn trailing_zero_count_u32(a: u32) -> u32

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Counts the number of trailing zero bits in a u32.

See in docs.rs: trailing_zero_count_u32.

pub fn trailing_zero_count_u64(a: u64) -> u64

Available on (crate features dep_safe_arch) and (target features bmi1) and (x86 or x86-64) only.

Counts the number of trailing zero bits in a u64.

See in docs.rs: trailing_zero_count_u64.

impl Arch

Functions requiring the bmi2 target feature.

---

See: https://en.wikipedia.org/wiki/X86_Bit_manipulation_instruction_set#BMI2_(Bit_Manipulation_Instruction_Set_2)

pub fn bit_zero_high_index_u32(a: u32, index: u32) -> u32

Available on (crate features dep_safe_arch) and (target features bmi2) and (x86 or x86-64) only.

Zero out all high bits in a u32 starting at the index given.

See in docs.rs: bit_zero_high_index_u32.

pub fn bit_zero_high_index_u64(a: u64, index: u32) -> u64

Available on (crate features dep_safe_arch) and (target features bmi2) and (x86 or x86-64) only.

Zero out all high bits in a u64 starting at the index given.

See in docs.rs: bit_zero_high_index_u64.

pub fn mul_extended_u32(a: u32, b: u32, extra: &mut u32) -> u32

Available on (crate features dep_safe_arch) and (target features bmi2) and (x86 or x86-64) only.

Multiply two u32, outputting the low bits and storing the high bits in the reference.

See in docs.rs: mul_extended_u32.

pub fn mul_extended_u64(a: u64, b: u64, extra: &mut u64) -> u64

Available on (crate features dep_safe_arch) and (target features bmi2) and (x86 or x86-64) only.

Multiply two u64, outputting the low bits and storing the high bits in the reference.

See in docs.rs: mul_extended_u64.

pub fn population_deposit_u32(a: u32, index: u32) -> u32

Available on (crate features dep_safe_arch) and (target features bmi2) and (x86 or x86-64) only.

Deposit contiguous low bits from a u32 according to a mask.

See in docs.rs: population_deposit_u32.

pub fn population_deposit_u64(a: u64, index: u64) -> u64

Available on (crate features dep_safe_arch) and (target features bmi2) and (x86 or x86-64) only.

Deposit contiguous low bits from a u64 according to a mask.

See in docs.rs: population_deposit_u64.

pub fn population_extract_u32(a: u32, index: u32) -> u32

Available on (crate features dep_safe_arch) and (target features bmi2) and (x86 or x86-64) only.

Extract bits from a u32 according to a mask.

See in docs.rs: population_extract_u32.

pub fn population_extract_u64(a: u64, index: u64) -> u64

Available on (crate features dep_safe_arch) and (target features bmi2) and (x86 or x86-64) only.

Extract bits from a u64 according to a mask.

See in docs.rs: population_extract_u64.

impl Arch

Functions requiring the fma target feature.

---

See: https://en.wikipedia.org/wiki/FMA_instruction_set

pub fn fused_mul_add_m128(a: m128, b: m128, c: m128) -> m128

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) + c

See in docs.rs: fused_mul_add_m128.

pub fn fused_mul_add_m128_s(a: m128, b: m128, c: m128) -> m128

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Low lane fused (a * b) + c, other lanes unchanged

See in docs.rs: fused_mul_add_m128_s.

pub fn fused_mul_add_m128d(a: m128d, b: m128d, c: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) + c

See in docs.rs: fused_mul_add_m128d.

pub fn fused_mul_add_m128d_s(a: m128d, b: m128d, c: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Low lane fused (a * b) + c, other lanes unchanged

See in docs.rs: fused_mul_add_m128d_s.

pub fn fused_mul_add_m256(a: m256, b: m256, c: m256) -> m256

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) + c

See in docs.rs: fused_mul_add_m256.

pub fn fused_mul_add_m256d(a: m256d, b: m256d, c: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) + c

See in docs.rs: fused_mul_add_m256d.

pub fn fused_mul_addsub_m128(a: m128, b: m128, c: m128) -> m128

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) addsub c (adds odd lanes and subtracts even lanes)

See in docs.rs: fused_mul_addsub_m128.

pub fn fused_mul_addsub_m128d(a: m128d, b: m128d, c: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) addsub c (adds odd lanes and subtracts even lanes)

See in docs.rs: fused_mul_addsub_m128d.

pub fn fused_mul_addsub_m256(a: m256, b: m256, c: m256) -> m256

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) addsub c (adds odd lanes and subtracts even lanes)

See in docs.rs: fused_mul_addsub_m256.

pub fn fused_mul_addsub_m256d(a: m256d, b: m256d, c: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) addsub c (adds odd lanes and subtracts even lanes)

See in docs.rs: fused_mul_addsub_m256d.

pub fn fused_mul_neg_add_m128(a: m128, b: m128, c: m128) -> m128

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused -(a * b) + c

See in docs.rs: fused_mul_neg_add_m128.

pub fn fused_mul_neg_add_m128_s(a: m128, b: m128, c: m128) -> m128

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Low lane -(a * b) + c, other lanes unchanged.

See in docs.rs: fused_mul_neg_add_m128_s.

pub fn fused_mul_neg_add_m128d(a: m128d, b: m128d, c: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused -(a * b) + c

See in docs.rs: fused_mul_neg_add_m128d.

pub fn fused_mul_neg_add_m128d_s(a: m128d, b: m128d, c: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Low lane -(a * b) + c, other lanes unchanged.

See in docs.rs: fused_mul_neg_add_m128d_s.

pub fn fused_mul_neg_add_m256(a: m256, b: m256, c: m256) -> m256

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused -(a * b) + c

See in docs.rs: fused_mul_neg_add_m256.

pub fn fused_mul_neg_add_m256d(a: m256d, b: m256d, c: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused -(a * b) + c

See in docs.rs: fused_mul_neg_add_m256d.

pub fn fused_mul_neg_sub_m128(a: m128, b: m128, c: m128) -> m128

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused -(a * b) - c

See in docs.rs: fused_mul_neg_sub_m128.

pub fn fused_mul_neg_sub_m128_s(a: m128, b: m128, c: m128) -> m128

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Low lane fused -(a * b) - c, other lanes unchanged.

See in docs.rs: fused_mul_neg_sub_m128_s.

pub fn fused_mul_neg_sub_m128d(a: m128d, b: m128d, c: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused -(a * b) - c

See in docs.rs: fused_mul_neg_sub_m128d.

pub fn fused_mul_neg_sub_m128d_s(a: m128d, b: m128d, c: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Low lane fused -(a * b) - c, other lanes unchanged.

See in docs.rs: fused_mul_neg_sub_m128d_s.

pub fn fused_mul_neg_sub_m256(a: m256, b: m256, c: m256) -> m256

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused -(a * b) - c

See in docs.rs: fused_mul_neg_sub_m256.

pub fn fused_mul_neg_sub_m256d(a: m256d, b: m256d, c: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused -(a * b) - c

See in docs.rs: fused_mul_neg_sub_m256d.

pub fn fused_mul_sub_m128(a: m128, b: m128, c: m128) -> m128

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) - c

See in docs.rs: fused_mul_sub_m128.

pub fn fused_mul_sub_m128_s(a: m128, b: m128, c: m128) -> m128

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Low lane fused (a * b) - c, other lanes unchanged.

See in docs.rs: fused_mul_sub_m128_s.

pub fn fused_mul_sub_m128d(a: m128d, b: m128d, c: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) - c

See in docs.rs: fused_mul_sub_m128d.

pub fn fused_mul_sub_m128d_s(a: m128d, b: m128d, c: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Low lane fused (a * b) - c, other lanes unchanged.

See in docs.rs: fused_mul_sub_m128d_s.

pub fn fused_mul_sub_m256(a: m256, b: m256, c: m256) -> m256

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) - c

See in docs.rs: fused_mul_sub_m256.

pub fn fused_mul_sub_m256d(a: m256d, b: m256d, c: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) - c

See in docs.rs: fused_mul_sub_m256d.

pub fn fused_mul_subadd_m128(a: m128, b: m128, c: m128) -> m128

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) subadd c (subtracts odd lanes and adds even lanes)

See in docs.rs: fused_mul_subadd_m128.

pub fn fused_mul_subadd_m128d(a: m128d, b: m128d, c: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) subadd c (subtracts odd lanes and adds even lanes)

See in docs.rs: fused_mul_subadd_m128d.

pub fn fused_mul_subadd_m256(a: m256, b: m256, c: m256) -> m256

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) subadd c (subtracts odd lanes and adds even lanes)

See in docs.rs: fused_mul_subadd_m256.

pub fn fused_mul_subadd_m256d(a: m256d, b: m256d, c: m256d) -> m256d

Available on (crate features dep_safe_arch) and (target features fma) and (x86 or x86-64) only.

Lanewise fused (a * b) subadd c (subtracts odd lanes and adds even lanes)

See in docs.rs: fused_mul_subadd_m256d.

impl Arch

Functions requiring the lzcnt target feature.

---

See: https://en.wikipedia.org/wiki/X86_Bit_manipulation_instruction_set#ABM_(Advanced_Bit_Manipulation)

pub fn leading_zero_count_u32(a: u32) -> u32

Available on (crate features dep_safe_arch) and (target features lzcnt) and (x86 or x86-64) only.

Count the leading zeroes in a u32.

See in docs.rs: leading_zero_count_u32.

pub fn leading_zero_count_u64(a: u64) -> u64

Available on (crate features dep_safe_arch) and (target features lzcnt) and (x86 or x86-64) only.

Count the leading zeroes in a u64.

See in docs.rs: leading_zero_count_u64.

impl Arch

Functions requiring the pclmulqdq target feature.

---

See: https://en.wikipedia.org/wiki/CLMUL_instruction_set

pub fn mul_i64_carryless_m128i<const IMM: i32>(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features pclmulqdq) and (x86 or x86-64) only.

Performs a “carryless” multiplication of two i64 values.

See in docs.rs: mul_i64_carryless_m128i.

impl Arch

Functions requiring the popcnt target feature.

---

See: https://en.wikipedia.org/wiki/X86_Bit_manipulation_instruction_set#ABM_(Advanced_Bit_Manipulation)

pub fn population_count_i32(a: i32) -> i32

Available on (crate features dep_safe_arch) and (target features popcnt) and (x86 or x86-64) only.

Count the number of bits set within an i32

See in docs.rs: population_count_i32.

pub fn population_count_i64(a: i64) -> i32

Available on (crate features dep_safe_arch) and (target features popcnt) and (x86 or x86-64) only.

Count the number of bits set within an i64

See in docs.rs: population_count_i64.

impl Arch

Functions requiring the rdrand target feature.

---

See: https://en.wikipedia.org/wiki/RDRAND

pub fn rdrand_u16(out: &mut u16) -> i32

Available on (crate features dep_safe_arch) and (target features rdrand) and (x86 or x86-64) only.

Try to obtain a random u16 from the hardware RNG.

See in docs.rs: rdrand_u16.

pub fn rdrand_u32(out: &mut u32) -> i32

Available on (crate features dep_safe_arch) and (target features rdrand) and (x86 or x86-64) only.

Try to obtain a random u32 from the hardware RNG.

See in docs.rs: rdrand_u32.

pub fn rdrand_u64(out: &mut u64) -> i32

Available on (crate features dep_safe_arch) and (target features rdrand) and (x86 or x86-64) only.

Try to obtain a random u64 from the hardware RNG.

See in docs.rs: rdrand_u64.

impl Arch

Functions requiring the rdseed target feature.

---

See: https://en.wikipedia.org/wiki/RDRAND

pub fn rdseed_u16(out: &mut u16) -> i32

Available on (crate features dep_safe_arch) and (target features rdseed) and (x86 or x86-64) only.

Try to obtain a random u16 from the hardware RNG.

See in docs.rs: rdseed_u16.

pub fn rdseed_u32(out: &mut u32) -> i32

Available on (crate features dep_safe_arch) and (target features rdseed) and (x86 or x86-64) only.

Try to obtain a random u32 from the hardware RNG.

See in docs.rs: rdseed_u32.

pub fn rdseed_u64(out: &mut u64) -> i32

Available on (crate features dep_safe_arch) and (target features rdseed) and (x86 or x86-64) only.

Try to obtain a random u64 from the hardware RNG.

See in docs.rs: rdseed_u64.

impl Arch

Functions requiring the sse target feature.

---

See: https://en.wikipedia.org/wiki/Streaming_SIMD_Extensions

pub fn add_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise a + b.

See in docs.rs: add_m128.

pub fn add_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane a + b, other lanes unchanged.

See in docs.rs: add_m128_s.

pub fn bitand_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Bitwise a & b.

See in docs.rs: bitand_m128.

pub fn bitandnot_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Bitwise (!a) & b.

See in docs.rs: bitandnot_m128.

pub fn bitor_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Bitwise a | b.

See in docs.rs: bitor_m128.

pub fn bitxor_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Bitwise a ^ b.

See in docs.rs: bitxor_m128.

pub fn cmp_eq_i32_m128_s(a: m128, b: m128) -> i32

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane equality.

See in docs.rs: cmp_eq_i32_m128_s.

pub fn cmp_eq_mask_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise a == b.

See in docs.rs: cmp_eq_mask_m128.

pub fn cmp_eq_mask_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane a == b, other lanes unchanged.

See in docs.rs: cmp_eq_mask_m128_s.

pub fn cmp_ge_i32_m128_s(a: m128, b: m128) -> i32

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane greater than or equal to.

See in docs.rs: cmp_ge_i32_m128_s.

pub fn cmp_ge_mask_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise a >= b.

See in docs.rs: cmp_ge_mask_m128.

pub fn cmp_ge_mask_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane a >= b, other lanes unchanged.

See in docs.rs: cmp_ge_mask_m128_s.

pub fn cmp_gt_i32_m128_s(a: m128, b: m128) -> i32

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane greater than.

See in docs.rs: cmp_gt_i32_m128_s.

pub fn cmp_gt_mask_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise a > b.

See in docs.rs: cmp_gt_mask_m128.

pub fn cmp_gt_mask_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane a > b, other lanes unchanged.

See in docs.rs: cmp_gt_mask_m128_s.

pub fn cmp_le_i32_m128_s(a: m128, b: m128) -> i32

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane less than or equal to.

See in docs.rs: cmp_le_i32_m128_s.

pub fn cmp_le_mask_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise a <= b.

See in docs.rs: cmp_le_mask_m128.

pub fn cmp_le_mask_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane a <= b, other lanes unchanged.

See in docs.rs: cmp_le_mask_m128_s.

pub fn cmp_lt_i32_m128_s(a: m128, b: m128) -> i32

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane less than.

See in docs.rs: cmp_lt_i32_m128_s.

pub fn cmp_lt_mask_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise a < b.

See in docs.rs: cmp_lt_mask_m128.

pub fn cmp_lt_mask_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane a < b, other lanes unchanged.

See in docs.rs: cmp_lt_mask_m128_s.

pub fn cmp_neq_i32_m128_s(a: m128, b: m128) -> i32

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane not equal to.

See in docs.rs: cmp_neq_i32_m128_s.

pub fn cmp_neq_mask_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise a != b.

See in docs.rs: cmp_neq_mask_m128.

pub fn cmp_neq_mask_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane a != b, other lanes unchanged.

See in docs.rs: cmp_neq_mask_m128_s.

pub fn cmp_nge_mask_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise !(a >= b).

See in docs.rs: cmp_nge_mask_m128.

pub fn cmp_nge_mask_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane !(a >= b), other lanes unchanged.

See in docs.rs: cmp_nge_mask_m128_s.

pub fn cmp_ngt_mask_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise !(a > b).

See in docs.rs: cmp_ngt_mask_m128.

pub fn cmp_ngt_mask_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane !(a > b), other lanes unchanged.

See in docs.rs: cmp_ngt_mask_m128_s.

pub fn cmp_nle_mask_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise !(a <= b).

See in docs.rs: cmp_nle_mask_m128.

pub fn cmp_nle_mask_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane !(a <= b), other lanes unchanged.

See in docs.rs: cmp_nle_mask_m128_s.

pub fn cmp_nlt_mask_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise !(a < b).

See in docs.rs: cmp_nlt_mask_m128.

pub fn cmp_nlt_mask_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane !(a < b), other lanes unchanged.

See in docs.rs: cmp_nlt_mask_m128_s.

pub fn cmp_ordered_mask_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise (!a.is_nan()) & (!b.is_nan()).

See in docs.rs: cmp_ordered_mask_m128.

pub fn cmp_ordered_mask_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane (!a.is_nan()) & (!b.is_nan()), other lanes unchanged.

See in docs.rs: cmp_ordered_mask_m128_s.

pub fn cmp_unord_mask_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise a.is_nan() | b.is_nan().

See in docs.rs: cmp_unord_mask_m128.

pub fn cmp_unord_mask_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane a.is_nan() | b.is_nan(), other lanes unchanged.

See in docs.rs: cmp_unord_mask_m128_s.

pub fn convert_i32_replace_m128_s(a: m128, i: i32) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Convert i32 to f32 and replace the low lane of the input.

See in docs.rs: convert_i32_replace_m128_s.

pub fn div_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise a / b.

See in docs.rs: div_m128.

pub fn div_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane a / b, other lanes unchanged.

See in docs.rs: div_m128_s.

pub fn get_f32_from_m128_s(a: m128) -> f32

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Gets the low lane as an individual f32 value.

See in docs.rs: get_f32_from_m128_s.

pub fn get_i32_from_m128_s(a: m128) -> i32

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Converts the low lane to i32 and extracts as an individual value.

See in docs.rs: get_i32_from_m128_s.

pub fn load_f32_m128_s(a: &f32) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Loads the f32 reference into the low lane of the register.

See in docs.rs: load_f32_m128_s.

pub fn load_f32_splat_m128(a: &f32) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Loads the f32 reference into all lanes of a register.

See in docs.rs: load_f32_splat_m128.

pub fn load_m128(a: &m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Loads the reference into a register.

See in docs.rs: load_m128.

pub fn load_reverse_m128(a: &m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Loads the reference into a register with reversed order.

See in docs.rs: load_reverse_m128.

pub fn load_unaligned_m128(a: &[f32; 4]) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Loads the reference into a register.

See in docs.rs: load_unaligned_m128.

pub fn max_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise max(a, b).

See in docs.rs: max_m128.

pub fn max_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane max(a, b), other lanes unchanged.

See in docs.rs: max_m128_s.

pub fn min_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise min(a, b).

See in docs.rs: min_m128.

pub fn min_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane min(a, b), other lanes unchanged.

See in docs.rs: min_m128_s.

pub fn move_high_low_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Move the high lanes of b to the low lanes of a, other lanes unchanged.

See in docs.rs: move_high_low_m128.

pub fn move_low_high_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Move the low lanes of b to the high lanes of a, other lanes unchanged.

See in docs.rs: move_low_high_m128.

pub fn move_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Move the low lane of b to a, other lanes unchanged.

See in docs.rs: move_m128_s.

pub fn move_mask_m128(a: m128) -> i32

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Gathers the sign bit of each lane.

See in docs.rs: move_mask_m128.

pub fn mul_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise a * b.

See in docs.rs: mul_m128.

pub fn mul_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane a * b, other lanes unchanged.

See in docs.rs: mul_m128_s.

pub fn reciprocal_m128(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise 1.0 / a approximation.

See in docs.rs: reciprocal_m128.

pub fn reciprocal_m128_s(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane 1.0 / a approximation, other lanes unchanged.

See in docs.rs: reciprocal_m128_s.

pub fn reciprocal_sqrt_m128(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise 1.0 / sqrt(a) approximation.

See in docs.rs: reciprocal_sqrt_m128.

pub fn reciprocal_sqrt_m128_s(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane 1.0 / sqrt(a) approximation, other lanes unchanged.

See in docs.rs: reciprocal_sqrt_m128_s.

pub fn set_m128(three: f32, two: f32, one: f32, zero: f32) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Sets the args into an m128, first arg is the high lane.

See in docs.rs: set_m128.

pub fn set_m128_s(low: f32) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Sets the args into an m128, first arg is the high lane.

See in docs.rs: set_m128_s.

pub fn set_reversed_m128(zero: f32, one: f32, two: f32, three: f32) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Sets the args into an m128, first arg is the low lane.

See in docs.rs: set_reversed_m128.

pub fn set_splat_m128(all: f32) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Splats the value to all lanes.

See in docs.rs: set_splat_m128.

pub fn shuffle_abi_f32_all_m128<const MASK: i32>(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Shuffle the f32 lanes from $a and $b together using an immediate control value.

See in docs.rs: shuffle_abi_f32_all_m128.

pub fn sqrt_m128(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise sqrt(a).

See in docs.rs: sqrt_m128.

pub fn sqrt_m128_s(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane sqrt(a), other lanes unchanged.

See in docs.rs: sqrt_m128_s.

pub fn store_m128(r: &mut m128, a: m128)

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Stores the value to the reference given.

See in docs.rs:store_m128.

pub fn store_m128_s(r: &mut f32, a: m128)

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Stores the low lane value to the reference given.

See in docs.rs:store_m128_s.

pub fn store_reverse_m128(r: &mut m128, a: m128)

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Stores the value to the reference given in reverse order.

See in docs.rs:store_reverse_m128.

pub fn store_splat_m128(r: &mut m128, a: m128)

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Stores the low lane value to all lanes of the reference given.

See in docs.rs:store_splat_m128.

pub fn store_unaligned_m128(r: &mut [f32; 4], a: m128)

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Stores the value to the reference given.

See in docs.rs:store_unaligned_m128.

pub fn sub_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Lanewise a - b.

See in docs.rs: sub_m128.

pub fn sub_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Low lane a - b, other lanes unchanged.

See in docs.rs: sub_m128_s.

pub fn transpose_four_m128( a: &mut m128, b: &mut m128, c: &mut m128, d: &mut m128, )

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Transpose four m128 as if they were a 4x4 matrix.

See in docs.rs:transpose_four_m128.

pub fn unpack_high_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Unpack and interleave high lanes of a and b.

See in docs.rs: unpack_high_m128.

pub fn unpack_low_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Unpack and interleave low lanes of a and b.

See in docs.rs: unpack_low_m128.

pub fn zeroed_m128() -> m128

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

All lanes zero.

See in docs.rs: zeroed_m128.

impl Arch

Generic functions requiring the sse target feature.

---

See: https://en.wikipedia.org/wiki/Streaming_SIMD_Extensions

pub fn prefetch_et0<T>(addr: &T)

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Fetches the cache line containing addr into all levels of the cache hierarchy, anticipating write.

pub fn prefetch_et1<T>(addr: &T)

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Fetches into L2 and higher, anticipating write.

pub fn prefetch_nta<T>(addr: &T)

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Fetch data using the non-temporal access (NTA) hint.

It may be a place closer than main memory but outside of the cache hierarchy.

This is used to reduce access latency without polluting the cache.

pub fn prefetch_t0<T>(addr: &T)

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Fetches the cache line containing addr into all levels of the cache hierarchy.

pub fn prefetch_t1<T>(addr: &T)

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Fetches into L2 and higher.

pub fn prefetch_t2<T>(addr: &T)

Available on (crate features dep_safe_arch) and (target features sse) and (x86 or x86-64) only.

Fetches into L3 and higher or an implementation-specific choice (e.g., L2 if there is no L3).

impl Arch

Functions requiring the sse2 target feature.

---

See: https://en.wikipedia.org/wiki/SSE2

pub fn add_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a + b with lanes as i16.

See in docs.rs: add_i16_m128i.

pub fn add_i32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a + b with lanes as i32.

See in docs.rs: add_i32_m128i.

pub fn add_i64_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a + b with lanes as i64.

See in docs.rs: add_i64_m128i.

pub fn add_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a + b with lanes as i8.

See in docs.rs: add_i8_m128i.

pub fn add_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a + b.

See in docs.rs: add_m128d.

pub fn add_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lowest lane a + b, high lane unchanged.

See in docs.rs: add_m128d_s.

pub fn add_saturating_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise saturating a + b with lanes as i16.

See in docs.rs: add_saturating_i16_m128i.

pub fn add_saturating_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise saturating a + b with lanes as i8.

See in docs.rs: add_saturating_i8_m128i.

pub fn add_saturating_u16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise saturating a + b with lanes as u16.

See in docs.rs: add_saturating_u16_m128i.

pub fn add_saturating_u8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise saturating a + b with lanes as u8.

See in docs.rs: add_saturating_u8_m128i.

pub fn average_u16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise average of the u16 values.

See in docs.rs: average_u16_m128i.

pub fn average_u8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise average of the u8 values.

See in docs.rs: average_u8_m128i.

pub fn bitand_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bitwise a & b.

See in docs.rs: bitand_m128d.

pub fn bitand_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bitwise a & b.

See in docs.rs: bitand_m128i.

pub fn bitandnot_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bitwise (!a) & b.

See in docs.rs: bitandnot_m128d.

pub fn bitandnot_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bitwise (!a) & b.

See in docs.rs: bitandnot_m128i.

pub fn bitor_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bitwise a | b.

See in docs.rs: bitor_m128d.

pub fn bitor_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bitwise a | b.

See in docs.rs: bitor_m128i.

pub fn bitxor_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bitwise a ^ b.

See in docs.rs: bitxor_m128d.

pub fn bitxor_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bitwise a ^ b.

See in docs.rs: bitxor_m128i.

pub fn byte_shl_imm_u128_m128i<const IMM: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shifts all bits in the entire register left by a number of bytes.

See in docs.rs: byte_shl_imm_u128_m128i.

pub fn byte_shr_imm_u128_m128i<const IMM: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shifts all bits in the entire register right by a number of bytes.

See in docs.rs: byte_shr_imm_u128_m128i.

pub fn cast_to_m128_from_m128d(a: m128d) -> m128

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bit-preserving cast to m128 from m128d

See in docs.rs: cast_to_m128_from_m128d.

pub fn cast_to_m128_from_m128i(a: m128i) -> m128

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bit-preserving cast to m128 from m128i

See in docs.rs: cast_to_m128_from_m128i.

pub fn cast_to_m128d_from_m128(a: m128) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bit-preserving cast to m128d from m128

See in docs.rs: cast_to_m128d_from_m128.

pub fn cast_to_m128d_from_m128i(a: m128i) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bit-preserving cast to m128d from m128i

See in docs.rs: cast_to_m128d_from_m128i.

pub fn cast_to_m128i_from_m128(a: m128) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bit-preserving cast to m128i from m128

See in docs.rs: cast_to_m128i_from_m128.

pub fn cast_to_m128i_from_m128d(a: m128d) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Bit-preserving cast to m128i from m128d

See in docs.rs: cast_to_m128i_from_m128d.

pub fn cmp_eq_i32_m128d_s(a: m128d, b: m128d) -> i32

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane f64 equal to.

See in docs.rs: cmp_eq_i32_m128d_s.

pub fn cmp_eq_mask_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a == b with lanes as i16.

See in docs.rs: cmp_eq_mask_i16_m128i.

pub fn cmp_eq_mask_i32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a == b with lanes as i32.

See in docs.rs: cmp_eq_mask_i32_m128i.

pub fn cmp_eq_mask_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a == b with lanes as i8.

See in docs.rs: cmp_eq_mask_i8_m128i.

pub fn cmp_eq_mask_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a == b, mask output.

See in docs.rs: cmp_eq_mask_m128d.

pub fn cmp_eq_mask_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane a == b, other lanes unchanged.

See in docs.rs: cmp_eq_mask_m128d_s.

pub fn cmp_ge_i32_m128d_s(a: m128d, b: m128d) -> i32

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane f64 greater than or equal to.

See in docs.rs: cmp_ge_i32_m128d_s.

pub fn cmp_ge_mask_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a >= b.

See in docs.rs: cmp_ge_mask_m128d.

pub fn cmp_ge_mask_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane a >= b, other lanes unchanged.

See in docs.rs: cmp_ge_mask_m128d_s.

pub fn cmp_gt_i32_m128d_s(a: m128d, b: m128d) -> i32

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane f64 greater than.

See in docs.rs: cmp_gt_i32_m128d_s.

pub fn cmp_gt_mask_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a > b with lanes as i16.

See in docs.rs: cmp_gt_mask_i16_m128i.

pub fn cmp_gt_mask_i32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a > b with lanes as i32.

See in docs.rs: cmp_gt_mask_i32_m128i.

pub fn cmp_gt_mask_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a > b with lanes as i8.

See in docs.rs: cmp_gt_mask_i8_m128i.

pub fn cmp_gt_mask_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a > b.

See in docs.rs: cmp_gt_mask_m128d.

pub fn cmp_gt_mask_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane a > b, other lanes unchanged.

See in docs.rs: cmp_gt_mask_m128d_s.

pub fn cmp_le_i32_m128d_s(a: m128d, b: m128d) -> i32

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane f64 less than or equal to.

See in docs.rs: cmp_le_i32_m128d_s.

pub fn cmp_le_mask_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a <= b.

See in docs.rs: cmp_le_mask_m128d.

pub fn cmp_le_mask_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane a <= b, other lanes unchanged.

See in docs.rs: cmp_le_mask_m128d_s.

pub fn cmp_lt_i32_m128d_s(a: m128d, b: m128d) -> i32

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane f64 less than.

See in docs.rs: cmp_lt_i32_m128d_s.

pub fn cmp_lt_mask_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a < b with lanes as i16.

See in docs.rs: cmp_lt_mask_i16_m128i.

pub fn cmp_lt_mask_i32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a < b with lanes as i32.

See in docs.rs: cmp_lt_mask_i32_m128i.

pub fn cmp_lt_mask_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a < b with lanes as i8.

See in docs.rs: cmp_lt_mask_i8_m128i.

pub fn cmp_lt_mask_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a < b.

See in docs.rs: cmp_lt_mask_m128d.

pub fn cmp_lt_mask_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane a < b, other lane unchanged.

See in docs.rs: cmp_lt_mask_m128d_s.

pub fn cmp_neq_i32_m128d_s(a: m128d, b: m128d) -> i32

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane f64 less than.

See in docs.rs: cmp_neq_i32_m128d_s.

pub fn cmp_neq_mask_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a != b.

See in docs.rs: cmp_neq_mask_m128d.

pub fn cmp_neq_mask_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane a != b, other lane unchanged.

See in docs.rs: cmp_neq_mask_m128d_s.

pub fn cmp_nge_mask_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise !(a >= b).

See in docs.rs: cmp_nge_mask_m128d.

pub fn cmp_nge_mask_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane !(a >= b), other lane unchanged.

See in docs.rs: cmp_nge_mask_m128d_s.

pub fn cmp_ngt_mask_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise !(a > b).

See in docs.rs: cmp_ngt_mask_m128d.

pub fn cmp_ngt_mask_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane !(a > b), other lane unchanged.

See in docs.rs: cmp_ngt_mask_m128d_s.

pub fn cmp_nle_mask_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise !(a <= b).

See in docs.rs: cmp_nle_mask_m128d.

pub fn cmp_nle_mask_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane !(a <= b), other lane unchanged.

See in docs.rs: cmp_nle_mask_m128d_s.

pub fn cmp_nlt_mask_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise !(a < b).

See in docs.rs: cmp_nlt_mask_m128d.

pub fn cmp_nlt_mask_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane !(a < b), other lane unchanged.

See in docs.rs: cmp_nlt_mask_m128d_s.

pub fn cmp_ordered_mask_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise (!a.is_nan()) & (!b.is_nan()).

See in docs.rs: cmp_ordered_mask_m128d.

pub fn cmp_ordered_mask_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane (!a.is_nan()) & (!b.is_nan()), other lane unchanged.

See in docs.rs: cmp_ordered_mask_m128d_s.

pub fn cmp_unord_mask_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a.is_nan() | b.is_nan().

See in docs.rs: cmp_unord_mask_m128d.

pub fn cmp_unord_mask_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane a.is_nan() | b.is_nan(), other lane unchanged.

See in docs.rs: cmp_unord_mask_m128d_s.

pub fn convert_i32_replace_m128d_s(a: m128d, i: i32) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Convert i32 to f64 and replace the low lane of the input.

See in docs.rs: convert_i32_replace_m128d_s.

pub fn convert_i64_replace_m128d_s(a: m128d, i: i64) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Convert i64 to f64 and replace the low lane of the input.

See in docs.rs: convert_i64_replace_m128d_s.

pub fn convert_m128_s_replace_m128d_s(a: m128d, b: m128) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Converts the lower f32 to f64 and replace the low lane of the input

See in docs.rs: convert_m128_s_replace_m128d_s.

pub fn convert_m128d_s_replace_m128_s(a: m128, b: m128d) -> m128

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Converts the low f64 to f32 and replaces the low lane of the input.

See in docs.rs: convert_m128d_s_replace_m128_s.

pub fn convert_to_i32_m128i_from_m128(a: m128) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Rounds the f32 lanes to i32 lanes.

See in docs.rs: convert_to_i32_m128i_from_m128.

pub fn convert_to_i32_m128i_from_m128d(a: m128d) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Rounds the two f64 lanes to the low two i32 lanes.

See in docs.rs: convert_to_i32_m128i_from_m128d.

pub fn convert_to_m128_from_i32_m128i(a: m128i) -> m128

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Rounds the four i32 lanes to four f32 lanes.

See in docs.rs: convert_to_m128_from_i32_m128i.

pub fn convert_to_m128_from_m128d(a: m128d) -> m128

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Rounds the two f64 lanes to the low two f32 lanes.

See in docs.rs: convert_to_m128_from_m128d.

pub fn convert_to_m128d_from_lower2_i32_m128i(a: m128i) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Rounds the lower two i32 lanes to two f64 lanes.

See in docs.rs: convert_to_m128d_from_lower2_i32_m128i.

pub fn convert_to_m128d_from_lower2_m128(a: m128) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Rounds the two f64 lanes to the low two f32 lanes.

See in docs.rs: convert_to_m128d_from_lower2_m128.

pub fn copy_i64_m128i_s(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Copy the low i64 lane to a new register, upper bits 0.

See in docs.rs: copy_i64_m128i_s.

pub fn copy_replace_low_f64_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Copies the a value and replaces the low lane with the low b value.

See in docs.rs: copy_replace_low_f64_m128d.

pub fn div_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a / b.

See in docs.rs: div_m128d.

pub fn div_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lowest lane a / b, high lane unchanged.

See in docs.rs: div_m128d_s.

pub fn extract_i16_as_i32_m128i<const LANE: i32>(a: m128i) -> i32

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Gets an i16 value out of an m128i, returns as i32.

See in docs.rs: extract_i16_as_i32_m128i.

pub fn get_f64_from_m128d_s(a: m128d) -> f64

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Gets the lower lane as an f64 value.

See in docs.rs: get_f64_from_m128d_s.

pub fn get_i32_from_m128d_s(a: m128d) -> i32

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Converts the lower lane to an i32 value.

See in docs.rs: get_i32_from_m128d_s.

pub fn get_i32_from_m128i_s(a: m128i) -> i32

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Converts the lower lane to an i32 value.

See in docs.rs: get_i32_from_m128i_s.

pub fn get_i64_from_m128d_s(a: m128d) -> i64

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Converts the lower lane to an i64 value.

See in docs.rs: get_i64_from_m128d_s.

pub fn get_i64_from_m128i_s(a: m128i) -> i64

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Converts the lower lane to an i64 value.

See in docs.rs: get_i64_from_m128i_s.

pub fn insert_i16_from_i32_m128i<const LANE: i32>(a: m128i, i: i32) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Inserts the low 16 bits of an i32 value into an m128i.

See in docs.rs: insert_i16_from_i32_m128i.

pub fn load_f64_m128d_s(a: &f64) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Loads the reference into the low lane of the register.

See in docs.rs: load_f64_m128d_s.

pub fn load_f64_splat_m128d(a: &f64) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Loads the f64 reference into all lanes of a register.

See in docs.rs: load_f64_splat_m128d.

pub fn load_i64_m128i_s(a: &m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Loads the low i64 into a register.

See in docs.rs: load_i64_m128i_s.

pub fn load_m128d(a: &m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Loads the reference into a register.

See in docs.rs: load_m128d.

pub fn load_m128i(a: &m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Loads the reference into a register.

See in docs.rs: load_m128i.

pub fn load_replace_high_m128d(a: m128d, b: &f64) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Loads the reference into a register, replacing the high lane.

See in docs.rs: load_replace_high_m128d.

pub fn load_replace_low_m128d(a: m128d, b: &f64) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Loads the reference into a register, replacing the low lane.

See in docs.rs: load_replace_low_m128d.

pub fn load_reverse_m128d(a: &m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Loads the reference into a register with reversed order.

See in docs.rs: load_reverse_m128d.

pub fn load_unaligned_m128d(a: &[f64; 2]) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Loads the reference into a register.

See in docs.rs: load_unaligned_m128d.

pub fn load_unaligned_m128i(a: &[u8; 16]) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Loads the reference into a register.

See in docs.rs: load_unaligned_m128i.

pub fn max_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise max(a, b) with lanes as i16.

See in docs.rs: max_i16_m128i.

pub fn max_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise max(a, b).

See in docs.rs: max_m128d.

pub fn max_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane max(a, b), other lanes unchanged.

See in docs.rs: max_m128d_s.

pub fn max_u8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise max(a, b) with lanes as u8.

See in docs.rs: max_u8_m128i.

pub fn min_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise min(a, b) with lanes as i16.

See in docs.rs: min_i16_m128i.

pub fn min_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise min(a, b).

See in docs.rs: min_m128d.

pub fn min_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane min(a, b), other lanes unchanged.

See in docs.rs: min_m128d_s.

pub fn min_u8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise min(a, b) with lanes as u8.

See in docs.rs: min_u8_m128i.

pub fn move_mask_i8_m128i(a: m128i) -> i32

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Gathers the i8 sign bit of each lane.

See in docs.rs: move_mask_i8_m128i.

pub fn move_mask_m128d(a: m128d) -> i32

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Gathers the sign bit of each lane.

See in docs.rs: move_mask_m128d.

pub fn mul_i16_horizontal_add_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Multiply i16 lanes producing i32 values, horizontal add pairs of i32 values to produce the final output.

See in docs.rs: mul_i16_horizontal_add_m128i.

pub fn mul_i16_keep_high_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a * b with lanes as i16, keep the high bits of the i32 intermediates.

See in docs.rs: mul_i16_keep_high_m128i.

pub fn mul_i16_keep_low_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a * b with lanes as i16, keep the low bits of the i32 intermediates.

See in docs.rs: mul_i16_keep_low_m128i.

pub fn mul_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a * b.

See in docs.rs: mul_m128d.

pub fn mul_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lowest lane a * b, high lane unchanged.

See in docs.rs: mul_m128d_s.

pub fn mul_u16_keep_high_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a * b with lanes as u16, keep the high bits of the u32 intermediates.

See in docs.rs: mul_u16_keep_high_m128i.

pub fn mul_widen_u32_odd_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Multiplies the odd u32 lanes and gives the widened (u64) results.

See in docs.rs: mul_widen_u32_odd_m128i.

pub fn pack_i16_to_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Saturating convert i16 to i8, and pack the values.

See in docs.rs: pack_i16_to_i8_m128i.

pub fn pack_i16_to_u8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Saturating convert i16 to u8, and pack the values.

See in docs.rs: pack_i16_to_u8_m128i.

pub fn pack_i32_to_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Saturating convert i32 to i16, and pack the values.

See in docs.rs: pack_i32_to_i16_m128i.

pub fn set_i16_m128i( a: i16, b: i16, c: i16, d: i16, e: i16, f: i16, g: i16, h: i16, ) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Sets the args into an m128i, first arg is the high lane.

See in docs.rs: set_i16_m128i.

pub fn set_i32_m128i(a: i32, b: i32, c: i32, d: i32) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Sets the args into an m128i, first arg is the high lane.

See in docs.rs: set_i32_m128i.

pub fn set_i32_m128i_s(i: i32) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Set an i32 as the low 32-bit lane of an m128i, other lanes blank.

See in docs.rs: set_i32_m128i_s.

pub fn set_i64_m128i(a: i64, b: i64) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Sets the args into an m128i, first arg is the high lane.

See in docs.rs: set_i64_m128i.

pub fn set_i64_m128i_s(i: i64) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Set an i64 as the low 64-bit lane of an m128i, other lanes blank.

See in docs.rs: set_i64_m128i_s.

pub fn set_i8_m128i( a: i8, b: i8, c: i8, d: i8, e: i8, f: i8, g: i8, h: i8, i: i8, j: i8, k: i8, l: i8, m: i8, n: i8, o: i8, p: i8, ) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Sets the args into an m128i, first arg is the high lane.

See in docs.rs: set_i8_m128i.

pub fn set_m128d(a: f64, b: f64) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Sets the args into an m128d, first arg is the high lane.

See in docs.rs: set_m128d.

pub fn set_m128d_s(a: f64) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Sets the args into the low lane of a m128d.

See in docs.rs: set_m128d_s.

pub fn set_reversed_i16_m128i( a: i16, b: i16, c: i16, d: i16, e: i16, f: i16, g: i16, h: i16, ) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Sets the args into an m128i, first arg is the low lane.

See in docs.rs: set_reversed_i16_m128i.

pub fn set_reversed_i32_m128i(a: i32, b: i32, c: i32, d: i32) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Sets the args into an m128i, first arg is the low lane.

See in docs.rs: set_reversed_i32_m128i.

pub fn set_reversed_i8_m128i( a: i8, b: i8, c: i8, d: i8, e: i8, f: i8, g: i8, h: i8, i: i8, j: i8, k: i8, l: i8, m: i8, n: i8, o: i8, p: i8, ) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Sets the args into an m128i, first arg is the low lane.

See in docs.rs: set_reversed_i8_m128i.

pub fn set_reversed_m128d(a: f64, b: f64) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Sets the args into an m128d, first arg is the low lane.

See in docs.rs: set_reversed_m128d.

pub fn set_splat_i16_m128i(i: i16) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Splats the i16 to all lanes of the m128i.

See in docs.rs: set_splat_i16_m128i.

pub fn set_splat_i32_m128i(i: i32) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Splats the i32 to all lanes of the m128i.

See in docs.rs: set_splat_i32_m128i.

pub fn set_splat_i64_m128i(i: i64) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Splats the i64 to both lanes of the m128i.

See in docs.rs: set_splat_i64_m128i.

pub fn set_splat_i8_m128i(i: i8) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Splats the i8 to all lanes of the m128i.

See in docs.rs: set_splat_i8_m128i.

pub fn set_splat_m128d(a: f64) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Splats the args into both lanes of the m128d.

See in docs.rs: set_splat_m128d.

pub fn shl_all_u16_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shift all u16 lanes to the left by the count in the lower u64 lane.

See in docs.rs: shl_all_u16_m128i.

pub fn shl_all_u32_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shift all u32 lanes to the left by the count in the lower u64 lane.

See in docs.rs: shl_all_u32_m128i.

pub fn shl_all_u64_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shift all u64 lanes to the left by the count in the lower u64 lane.

See in docs.rs: shl_all_u64_m128i.

pub fn shl_imm_u16_m128i<const IMM: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shifts all u16 lanes left by an immediate.

See in docs.rs: shl_imm_u16_m128i.

pub fn shl_imm_u32_m128i<const IMM: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shifts all u32 lanes left by an immediate.

See in docs.rs: shl_imm_u32_m128i.

pub fn shl_imm_u64_m128i<const IMM: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shifts both u64 lanes left by an immediate.

See in docs.rs: shl_imm_u64_m128i.

pub fn shr_all_i16_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shift each i16 lane to the right by the count in the lower i64 lane.

See in docs.rs: shr_all_i16_m128i.

pub fn shr_all_i32_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shift each i32 lane to the right by the count in the lower i64 lane.

See in docs.rs: shr_all_i32_m128i.

pub fn shr_all_u16_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shift each u16 lane to the right by the count in the lower u64 lane.

See in docs.rs: shr_all_u16_m128i.

pub fn shr_all_u32_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shift each u32 lane to the right by the count in the lower u64 lane.

See in docs.rs: shr_all_u32_m128i.

pub fn shr_all_u64_m128i(a: m128i, count: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shift each u64 lane to the right by the count in the lower u64 lane.

See in docs.rs: shr_all_u64_m128i.

pub fn shr_imm_i16_m128i<const IMM: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shifts all i16 lanes right by an immediate.

See in docs.rs: shr_imm_i16_m128i.

pub fn shr_imm_i32_m128i<const IMM: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shifts all i32 lanes right by an immediate.

See in docs.rs: shr_imm_i32_m128i.

pub fn shr_imm_u16_m128i<const IMM: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shifts all u16 lanes right by an immediate.

See in docs.rs: shr_imm_u16_m128i.

pub fn shr_imm_u32_m128i<const IMM: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shifts all u32 lanes right by an immediate.

See in docs.rs: shr_imm_u32_m128i.

pub fn shr_imm_u64_m128i<const IMM: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shifts both u64 lanes right by an immediate.

See in docs.rs: shr_imm_u64_m128i.

pub fn shuffle_abi_f64_all_m128d<const MASK: i32>(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shuffle the f64 lanes from $a and $b together using an immediate control value.

See in docs.rs: shuffle_abi_f64_all_m128d.

pub fn shuffle_ai_f32_all_m128i<const MASK: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shuffle the i32 lanes in $a using an immediate control value.

See in docs.rs: shuffle_ai_f32_all_m128i.

pub fn shuffle_ai_i16_h64all_m128i<const MASK: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shuffle the high i16 lanes in $a using an immediate control value.

See in docs.rs: shuffle_ai_i16_h64all_m128i.

pub fn shuffle_ai_i16_l64all_m128i<const MASK: i32>(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Shuffle the low i16 lanes in $a using an immediate control value.

See in docs.rs: shuffle_ai_i16_l64all_m128i.

pub fn sqrt_m128d(a: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise sqrt(a).

See in docs.rs: sqrt_m128d.

pub fn sqrt_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Low lane sqrt(b), upper lane is unchanged from a.

See in docs.rs: sqrt_m128d_s.

pub fn store_high_m128d_s(r: &mut f64, a: m128d)

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Stores the high lane value to the reference given.

See in docs.rs:store_high_m128d_s.

pub fn store_i64_m128i_s(r: &mut i64, a: m128i)

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Stores the value to the reference given.

See in docs.rs:store_i64_m128i_s.

pub fn store_m128d(r: &mut m128d, a: m128d)

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Stores the value to the reference given.

See in docs.rs:store_m128d.

pub fn store_m128d_s(r: &mut f64, a: m128d)

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Stores the low lane value to the reference given.

See in docs.rs:store_m128d_s.

pub fn store_m128i(r: &mut m128i, a: m128i)

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Stores the value to the reference given.

See in docs.rs:store_m128i.

pub fn store_reversed_m128d(r: &mut m128d, a: m128d)

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Stores the value to the reference given.

See in docs.rs:store_reversed_m128d.

pub fn store_splat_m128d(r: &mut m128d, a: m128d)

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Stores the low lane value to all lanes of the reference given.

See in docs.rs:store_splat_m128d.

pub fn store_unaligned_m128d(r: &mut [f64; 2], a: m128d)

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Stores the value to the reference given.

See in docs.rs:store_unaligned_m128d.

pub fn store_unaligned_m128i(r: &mut [u8; 16], a: m128i)

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Stores the value to the reference given.

See in docs.rs:store_unaligned_m128i.

pub fn sub_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a - b with lanes as i16.

See in docs.rs: sub_i16_m128i.

pub fn sub_i32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a - b with lanes as i32.

See in docs.rs: sub_i32_m128i.

pub fn sub_i64_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a - b with lanes as i64.

See in docs.rs: sub_i64_m128i.

pub fn sub_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a - b with lanes as i8.

See in docs.rs: sub_i8_m128i.

pub fn sub_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise a - b.

See in docs.rs: sub_m128d.

pub fn sub_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lowest lane a - b, high lane unchanged.

See in docs.rs: sub_m128d_s.

pub fn sub_saturating_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise saturating a - b with lanes as i16.

See in docs.rs: sub_saturating_i16_m128i.

pub fn sub_saturating_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise saturating a - b with lanes as i8.

See in docs.rs: sub_saturating_i8_m128i.

pub fn sub_saturating_u16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise saturating a - b with lanes as u16.

See in docs.rs: sub_saturating_u16_m128i.

pub fn sub_saturating_u8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Lanewise saturating a - b with lanes as u8.

See in docs.rs: sub_saturating_u8_m128i.

pub fn sum_of_u8_abs_diff_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Compute “sum of u8 absolute differences”.

See in docs.rs: sum_of_u8_abs_diff_m128i.

pub fn truncate_m128_to_m128i(a: m128) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Truncate the f32 lanes to i32 lanes.

See in docs.rs: truncate_m128_to_m128i.

pub fn truncate_m128d_to_m128i(a: m128d) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Truncate the f64 lanes to the lower i32 lanes (upper i32 lanes 0).

See in docs.rs: truncate_m128d_to_m128i.

pub fn truncate_to_i32_m128d_s(a: m128d) -> i32

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Truncate the lower lane into an i32.

See in docs.rs: truncate_to_i32_m128d_s.

pub fn truncate_to_i64_m128d_s(a: m128d) -> i64

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Truncate the lower lane into an i64.

See in docs.rs: truncate_to_i64_m128d_s.

pub fn unpack_high_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Unpack and interleave high i16 lanes of a and b.

See in docs.rs: unpack_high_i16_m128i.

pub fn unpack_high_i32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Unpack and interleave high i32 lanes of a and b.

See in docs.rs: unpack_high_i32_m128i.

pub fn unpack_high_i64_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Unpack and interleave high i64 lanes of a and b.

See in docs.rs: unpack_high_i64_m128i.

pub fn unpack_high_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Unpack and interleave high i8 lanes of a and b.

See in docs.rs: unpack_high_i8_m128i.

pub fn unpack_high_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Unpack and interleave high lanes of a and b.

See in docs.rs: unpack_high_m128d.

pub fn unpack_low_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Unpack and interleave low i16 lanes of a and b.

See in docs.rs: unpack_low_i16_m128i.

pub fn unpack_low_i32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Unpack and interleave low i32 lanes of a and b.

See in docs.rs: unpack_low_i32_m128i.

pub fn unpack_low_i64_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Unpack and interleave low i64 lanes of a and b.

See in docs.rs: unpack_low_i64_m128i.

pub fn unpack_low_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Unpack and interleave low i8 lanes of a and b.

See in docs.rs: unpack_low_i8_m128i.

pub fn unpack_low_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Unpack and interleave low lanes of a and b.

See in docs.rs: unpack_low_m128d.

pub fn zeroed_m128d() -> m128d

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

Both lanes zero.

See in docs.rs: zeroed_m128d.

pub fn zeroed_m128i() -> m128i

Available on (crate features dep_safe_arch) and (target features sse2) and (x86 or x86-64) only.

All lanes zero.

See in docs.rs: zeroed_m128i.

impl Arch

Functions requiring the sse3 target feature.

---

See: https://en.wikipedia.org/wiki/SSE3

pub fn add_horizontal_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse3) and (x86 or x86-64) only.

Add each lane horizontally, pack the outputs as a then b.

See in docs.rs: add_horizontal_m128.

pub fn add_horizontal_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse3) and (x86 or x86-64) only.

Add each lane horizontally, pack the outputs as a then b.

See in docs.rs: add_horizontal_m128d.

pub fn addsub_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse3) and (x86 or x86-64) only.

Alternately, from the top, add a lane and then subtract a lane.

See in docs.rs: addsub_m128.

pub fn addsub_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse3) and (x86 or x86-64) only.

Add the high lane and subtract the low lane.

See in docs.rs: addsub_m128d.

pub fn duplicate_even_lanes_m128(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse3) and (x86 or x86-64) only.

Duplicate the odd lanes to the even lanes.

See in docs.rs: duplicate_even_lanes_m128.

pub fn duplicate_low_lane_m128d_s(a: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse3) and (x86 or x86-64) only.

Copy the low lane of the input to both lanes of the output.

See in docs.rs: duplicate_low_lane_m128d_s.

pub fn duplicate_odd_lanes_m128(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse3) and (x86 or x86-64) only.

Duplicate the odd lanes to the even lanes.

See in docs.rs: duplicate_odd_lanes_m128.

pub fn sub_horizontal_m128(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse3) and (x86 or x86-64) only.

Subtract each lane horizontally, pack the outputs as a then b.

See in docs.rs: sub_horizontal_m128.

pub fn sub_horizontal_m128d(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse3) and (x86 or x86-64) only.

Subtract each lane horizontally, pack the outputs as a then b.

See in docs.rs: sub_horizontal_m128d.

impl Arch

Functions requiring the sse4.1 target feature.

---

See: https://en.wikipedia.org/wiki/SSE4#SSE4.1

pub fn blend_imm_i16_m128i<const IMM: i32>(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Blends the i16 lanes according to the immediate mask.

See in docs.rs: blend_imm_i16_m128i.

pub fn blend_imm_m128<const IMM: i32>(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Blends the lanes according to the immediate mask.

See in docs.rs: blend_imm_m128.

pub fn blend_imm_m128d<const IMM: i32>(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Blends the i16 lanes according to the immediate mask.

See in docs.rs: blend_imm_m128d.

pub fn blend_varying_i8_m128i(a: m128i, b: m128i, mask: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Blend the i8 lanes according to a runtime varying mask.

See in docs.rs: blend_varying_i8_m128i.

pub fn blend_varying_m128(a: m128, b: m128, mask: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Blend the lanes according to a runtime varying mask.

See in docs.rs: blend_varying_m128.

pub fn blend_varying_m128d(a: m128d, b: m128d, mask: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Blend the lanes according to a runtime varying mask.

See in docs.rs: blend_varying_m128d.

pub fn ceil_m128(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Round each lane to a whole number, towards positive infinity.

See in docs.rs: ceil_m128.

pub fn ceil_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Round the low lane of b toward positive infinity, other lanes a.

See in docs.rs: ceil_m128_s.

pub fn ceil_m128d(a: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Round each lane to a whole number, towards positive infinity.

See in docs.rs: ceil_m128d.

pub fn ceil_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Round the low lane of b toward positive infinity, high lane is a.

See in docs.rs: ceil_m128d_s.

pub fn cmp_eq_mask_i64_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Lanewise a == b with lanes as i64.

See in docs.rs: cmp_eq_mask_i64_m128i.

pub fn convert_to_i16_m128i_from_lower2_i16_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Convert the lower two i64 lanes to two i32 lanes.

See in docs.rs: convert_to_i16_m128i_from_lower2_i16_m128i.

pub fn convert_to_i16_m128i_from_lower8_i8_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Convert the lower eight i8 lanes to eight i16 lanes.

See in docs.rs: convert_to_i16_m128i_from_lower8_i8_m128i.

pub fn convert_to_i32_m128i_from_lower4_i16_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Convert the lower four i16 lanes to four i32 lanes.

See in docs.rs: convert_to_i32_m128i_from_lower4_i16_m128i.

pub fn convert_to_i32_m128i_from_lower4_i8_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Convert the lower four i8 lanes to four i32 lanes.

See in docs.rs: convert_to_i32_m128i_from_lower4_i8_m128i.

pub fn convert_to_i64_m128i_from_lower2_i32_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Convert the lower two i32 lanes to two i64 lanes.

See in docs.rs: convert_to_i64_m128i_from_lower2_i32_m128i.

pub fn convert_to_i64_m128i_from_lower2_i8_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Convert the lower two i8 lanes to two i64 lanes.

See in docs.rs: convert_to_i64_m128i_from_lower2_i8_m128i.

pub fn convert_to_u16_m128i_from_lower8_u8_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Convert the lower eight u8 lanes to eight u16 lanes.

See in docs.rs: convert_to_u16_m128i_from_lower8_u8_m128i.

pub fn convert_to_u32_m128i_from_lower4_u16_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Convert the lower four u16 lanes to four u32 lanes.

See in docs.rs: convert_to_u32_m128i_from_lower4_u16_m128i.

pub fn convert_to_u32_m128i_from_lower4_u8_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Convert the lower four u8 lanes to four u32 lanes.

See in docs.rs: convert_to_u32_m128i_from_lower4_u8_m128i.

pub fn convert_to_u64_m128i_from_lower2_u16_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Convert the lower two u16 lanes to two u64 lanes.

See in docs.rs: convert_to_u64_m128i_from_lower2_u16_m128i.

pub fn convert_to_u64_m128i_from_lower2_u32_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Convert the lower two u32 lanes to two u64 lanes.

See in docs.rs: convert_to_u64_m128i_from_lower2_u32_m128i.

pub fn convert_to_u64_m128i_from_lower2_u8_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Convert the lower two u8 lanes to two u64 lanes.

See in docs.rs: convert_to_u64_m128i_from_lower2_u8_m128i.

pub fn dot_product_m128<const IMM: i32>(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Performs a dot product of two m128 registers.

See in docs.rs: dot_product_m128.

pub fn dot_product_m128d<const IMM: i32>(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Performs a dot product of two m128d registers.

See in docs.rs: dot_product_m128d.

pub fn extract_f32_as_i32_bits_imm_m128<const IMM: i32>(a: m128) -> i32

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Gets the f32 lane requested. Returns as an i32 bit pattern.

See in docs.rs: extract_f32_as_i32_bits_imm_m128.

pub fn extract_i32_imm_m128i<const IMM: i32>(a: m128i) -> i32

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Gets the i32 lane requested. Only the lowest 2 bits are considered.

See in docs.rs: extract_i32_imm_m128i.

pub fn extract_i64_imm_m128i<const IMM: i32>(a: m128i) -> i64

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Gets the i64 lane requested. Only the lowest bit is considered.

See in docs.rs: extract_i64_imm_m128i.

pub fn extract_i8_as_i32_imm_m128i<const IMM: i32>(a: m128i) -> i32

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Gets the i8 lane requested. Only the lowest 4 bits are considered.

See in docs.rs: extract_i8_as_i32_imm_m128i.

pub fn floor_m128(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Round each lane to a whole number, towards negative infinity

See in docs.rs: floor_m128.

pub fn floor_m128_s(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Round the low lane of b toward negative infinity, other lanes a.

See in docs.rs: floor_m128_s.

pub fn floor_m128d(a: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Round each lane to a whole number, towards negative infinity

See in docs.rs: floor_m128d.

pub fn floor_m128d_s(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Round the low lane of b toward negative infinity, high lane is a.

See in docs.rs: floor_m128d_s.

pub fn insert_f32_imm_m128<const IMM: i32>(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Inserts a lane from $b into $a, optionally at a new position.

See in docs.rs: insert_f32_imm_m128.

pub fn insert_i32_imm_m128i<const IMM: i32>(a: m128i, new: i32) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Inserts a new value for the i32 lane specified.

See in docs.rs: insert_i32_imm_m128i.

pub fn insert_i64_imm_m128i<const IMM: i32>(a: m128i, new: i64) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Inserts a new value for the i64 lane specified.

See in docs.rs: insert_i64_imm_m128i.

pub fn insert_i8_imm_m128i<const IMM: i32>(a: m128i, new: i32) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Inserts a new value for the i64 lane specified.

See in docs.rs: insert_i8_imm_m128i.

pub fn max_i32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Lanewise max(a, b) with lanes as i32.

See in docs.rs: max_i32_m128i.

pub fn max_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Lanewise max(a, b) with lanes as i8.

See in docs.rs: max_i8_m128i.

pub fn max_u16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Lanewise max(a, b) with lanes as u16.

See in docs.rs: max_u16_m128i.

pub fn max_u32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Lanewise max(a, b) with lanes as u32.

See in docs.rs: max_u32_m128i.

pub fn min_i32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Lanewise min(a, b) with lanes as i32.

See in docs.rs: min_i32_m128i.

pub fn min_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Lanewise min(a, b) with lanes as i8.

See in docs.rs: min_i8_m128i.

pub fn min_position_u16_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Min u16 value, position, and other lanes zeroed.

See in docs.rs: min_position_u16_m128i.

pub fn min_u16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Lanewise min(a, b) with lanes as u16.

See in docs.rs: min_u16_m128i.

pub fn min_u32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Lanewise min(a, b) with lanes as u32.

See in docs.rs: min_u32_m128i.

pub fn mul_32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Lanewise a * b with 32-bit lanes.

See in docs.rs: mul_32_m128i.

pub fn mul_widen_i32_odd_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Multiplies the odd i32 lanes and gives the widened (i64) results.

See in docs.rs: mul_widen_i32_odd_m128i.

pub fn multi_packed_sum_abs_diff_u8_m128i<const IMM: i32>( a: m128i, b: m128i, ) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Computes eight u16 “sum of absolute difference” values according to the bytes selected.

See in docs.rs: multi_packed_sum_abs_diff_u8_m128i.

pub fn pack_i32_to_u16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Saturating convert i32 to u16, and pack the values.

See in docs.rs: pack_i32_to_u16_m128i.

pub fn round_m128<const MODE: i32>(a: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Rounds each lane in the style specified.

See in docs.rs: round_m128.

pub fn round_m128_s<const MODE: i32>(a: m128, b: m128) -> m128

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Rounds $b low as specified, other lanes use $a.

See in docs.rs: round_m128_s.

pub fn round_m128d<const MODE: i32>(a: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Rounds each lane in the style specified.

See in docs.rs: round_m128d.

pub fn round_m128d_s<const MODE: i32>(a: m128d, b: m128d) -> m128d

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Rounds $b low as specified, keeps $a high.

See in docs.rs: round_m128d_s.

pub fn test_all_ones_m128i(a: m128i) -> i32

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Tests if all bits are 1.

See in docs.rs: test_all_ones_m128i.

pub fn test_all_zeroes_m128i(a: m128i, mask: m128i) -> i32

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Returns if all masked bits are 0, (a & mask) as u128 == 0

See in docs.rs: test_all_zeroes_m128i.

pub fn test_mixed_ones_and_zeroes_m128i(a: m128i, mask: m128i) -> i32

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Returns if, among the masked bits, there’s both 0s and 1s

See in docs.rs: test_mixed_ones_and_zeroes_m128i.

pub fn testc_m128i(a: m128i, b: m128i) -> i32

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Compute the bitwise NOT of a and then AND with b, returns 1 if the result is zero, otherwise 0.

See in docs.rs: testc_m128i.

pub fn testz_m128i(a: m128i, b: m128i) -> i32

Available on (crate features dep_safe_arch) and (target features sse4.1) and (x86 or x86-64) only.

Computes the bitwise AND of 256 bits in a and b, returns 1 if the result is zero, otherwise 0.

See in docs.rs: testz_m128i.

impl Arch

Functions requiring the sse4.2 target feature.

---

See: https://en.wikipedia.org/wiki/SSE4#SSE4.2

pub fn cmp_gt_mask_i64_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.2) and (x86 or x86-64) only.

Lanewise a > b with lanes as i64.

See in docs.rs: cmp_gt_mask_i64_m128i.

pub fn crc32_u16(crc: u32, v: u16) -> u32

Available on (crate features dep_safe_arch) and (target features sse4.2) and (x86 or x86-64) only.

Accumulates the u16 into a running CRC32 value.

See in docs.rs: crc32_u16.

pub fn crc32_u32(crc: u32, v: u32) -> u32

Available on (crate features dep_safe_arch) and (target features sse4.2) and (x86 or x86-64) only.

Accumulates the u32 into a running CRC32 value.

See in docs.rs: crc32_u32.

pub fn crc32_u64(crc: u64, v: u64) -> u64

Available on (crate features dep_safe_arch) and (target features sse4.2) and (x86 or x86-64) only.

Accumulates the u64 into a running CRC32 value.

See in docs.rs: crc32_u64.

pub fn crc32_u8(crc: u32, v: u8) -> u32

Available on (crate features dep_safe_arch) and (target features sse4.2) and (x86 or x86-64) only.

Accumulates the u8 into a running CRC32 value.

See in docs.rs: crc32_u8.

pub fn search_explicit_str_for_index<const IMM: i32>( needle: m128i, needle_len: i32, haystack: m128i, haystack_len: i32, ) -> i32

Available on (crate features dep_safe_arch) and (target features sse4.2) and (x86 or x86-64) only.

Search for needle in `haystack, with explicit string length.

See in docs.rs: search_explicit_str_for_index.

pub fn search_explicit_str_for_mask<const IMM: i32>( needle: m128i, needle_len: i32, haystack: m128i, haystack_len: i32, ) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.2) and (x86 or x86-64) only.

Search for needle in `haystack, with explicit string length.

See in docs.rs: search_explicit_str_for_mask.

pub fn search_implicit_str_for_index<const IMM: i32>( needle: m128i, haystack: m128i, ) -> i32

Available on (crate features dep_safe_arch) and (target features sse4.2) and (x86 or x86-64) only.

Search for needle in `haystack, with implicit string length.

See in docs.rs: search_implicit_str_for_index.

pub fn search_implicit_str_for_mask<const IMM: i32>( needle: m128i, haystack: m128i, ) -> m128i

Available on (crate features dep_safe_arch) and (target features sse4.2) and (x86 or x86-64) only.

Search for needle in `haystack, with implicit string length.

See in docs.rs: search_implicit_str_for_mask.

impl Arch

Functions requiring the ssse3 target feature.

---

See: https://en.wikipedia.org/wiki/SSSE3

pub fn abs_i16_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Lanewise absolute value with lanes as i16.

See in docs.rs: abs_i16_m128i.

pub fn abs_i32_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Lanewise absolute value with lanes as i32.

See in docs.rs: abs_i32_m128i.

pub fn abs_i8_m128i(a: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Lanewise absolute value with lanes as i8.

See in docs.rs: abs_i8_m128i.

pub fn add_horizontal_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Add horizontal pairs of i16 values, pack the outputs as a then b.

See in docs.rs: add_horizontal_i16_m128i.

pub fn add_horizontal_i32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Add horizontal pairs of i32 values, pack the outputs as a then b.

See in docs.rs: add_horizontal_i32_m128i.

pub fn add_horizontal_saturating_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Add horizontal pairs of i16 values, saturating, pack the outputs as a then b.

See in docs.rs: add_horizontal_saturating_i16_m128i.

pub fn combined_byte_shr_imm_m128i<const IMM: i32>(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Counts $a as the high bytes and $b as the low bytes then performs a byte shift to the right by the immediate value.

See in docs.rs: combined_byte_shr_imm_m128i.

pub fn mul_i16_scale_round_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Multiply i16 lanes into i32 intermediates, keep the high 18 bits, round by adding 1, right shift by 1.

See in docs.rs: mul_i16_scale_round_m128i.

pub fn mul_u8i8_add_horizontal_saturating_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

This is dumb and weird.

See in docs.rs: mul_u8i8_add_horizontal_saturating_m128i.

pub fn shuffle_av_i8z_all_m128i(a: m128i, v: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Shuffle i8 lanes in a using i8 values in v.

See in docs.rs: shuffle_av_i8z_all_m128i.

pub fn sign_apply_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Applies the sign of i16 values in b to the values in a.

See in docs.rs: sign_apply_i16_m128i.

pub fn sign_apply_i32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Applies the sign of i32 values in b to the values in a.

See in docs.rs: sign_apply_i32_m128i.

pub fn sign_apply_i8_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Applies the sign of i8 values in b to the values in a.

See in docs.rs: sign_apply_i8_m128i.

pub fn sub_horizontal_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Subtract horizontal pairs of i16 values, pack the outputs as a then b.

See in docs.rs: sub_horizontal_i16_m128i.

pub fn sub_horizontal_i32_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Subtract horizontal pairs of i32 values, pack the outputs as a then b.

See in docs.rs: sub_horizontal_i32_m128i.

pub fn sub_horizontal_saturating_i16_m128i(a: m128i, b: m128i) -> m128i

Available on (crate features dep_safe_arch) and (target features ssse3) and (x86 or x86-64) only.

Subtract horizontal pairs of i16 values, saturating, pack the outputs as a then b.

See in docs.rs: sub_horizontal_saturating_i16_m128i.

Trait Implementations

impl Debug for Arch

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.