Trait subtle::ConstantTimeGreater

pub trait ConstantTimeGreater {
    fn ct_gt(&self, other: &Self) -> Choice;
}

A type which can be compared in some manner and be determined to be greater than another of the same type.

Required methods

fn ct_gt(&self, other: &Self) -> Choice

Determine whether self > other.

The bitwise-NOT of the return value of this function should be usable to determine if self <= other.

This function should execute in constant time.

Returns

A Choice with a set bit if self > other, and with no set bits otherwise.

Example

use subtle::ConstantTimeGreater;

let x: u8 = 13;
let y: u8 = 42;

let x_gt_y = x.ct_gt(&y);

assert_eq!(x_gt_y.unwrap_u8(), 0);

let y_gt_x = y.ct_gt(&x);

assert_eq!(y_gt_x.unwrap_u8(), 1);

let x_gt_x = x.ct_gt(&x);

assert_eq!(x_gt_x.unwrap_u8(), 0);

Loading content...

Implementations on Foreign Types

impl ConstantTimeGreater for u8

fn ct_gt(&self, other: &u8) -> Choice

Returns Choice::from(1) iff x > y, and Choice::from(0) iff x <= y.

Note

This algoritm would also work for signed integers if we first flip the top bit, e.g. let x: u8 = x ^ 0x80, etc.

impl ConstantTimeGreater for u16

fn ct_gt(&self, other: &u16) -> Choice

Returns Choice::from(1) iff x > y, and Choice::from(0) iff x <= y.

Note

This algoritm would also work for signed integers if we first flip the top bit, e.g. let x: u8 = x ^ 0x80, etc.

impl ConstantTimeGreater for u32

fn ct_gt(&self, other: &u32) -> Choice

Returns Choice::from(1) iff x > y, and Choice::from(0) iff x <= y.

Note

This algoritm would also work for signed integers if we first flip the top bit, e.g. let x: u8 = x ^ 0x80, etc.

impl ConstantTimeGreater for u64

fn ct_gt(&self, other: &u64) -> Choice

Returns Choice::from(1) iff x > y, and Choice::from(0) iff x <= y.

Note

This algoritm would also work for signed integers if we first flip the top bit, e.g. let x: u8 = x ^ 0x80, etc.

impl ConstantTimeGreater for u128

fn ct_gt(&self, other: &u128) -> Choice

Returns Choice::from(1) iff x > y, and Choice::from(0) iff x <= y.

Note

This algoritm would also work for signed integers if we first flip the top bit, e.g. let x: u8 = x ^ 0x80, etc.

Loading content...

Implementors

Loading content...