Struct secmem_alloc::zeroize::X86_64Sse2Zeroizer

pub struct X86_64Sse2Zeroizer;

This zeroizer uses inline asm with sse2 instructions if the pointer is 16 byte aligned, and otherwise uses VolatileWrite8Zeroizer. This zeroization technique is available for x86_64 platforms with sse2 cpu support on stable, and reasonably fast for 16 byte aligned pointers.

In addition to the volatile write we place a compiler fence right next to the volatile write. This should not be necessary for secure zeroization since the volatile semantics guarenties our writes are not elided, and they can not be delayed since we are deallocating the memory after zeroization. The use of this fence is therefore only a precaution.

This zeroization method can benefit (in terms of performance) from using the MemZeroizer::zeroize_mem_blocks function instead of MemZeroizer::zeroize_mem function if a minimum alignment is known at compile time.

Trait Implementations

impl Clone for X86_64Sse2Zeroizer

fn clone(&self) -> X86_64Sse2Zeroizer

Returns a copy of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for X86_64Sse2Zeroizer

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for X86_64Sse2Zeroizer

fn default() -> X86_64Sse2Zeroizer

Returns the “default value” for a type.

impl MemZeroizer for X86_64Sse2Zeroizer

unsafe fn zeroize_mem_blocks<const A: u8, const B: u8>(
    &self,
    ptr: *mut u8,
    len: usize
)

Zeroize the memory pointed to by ptr and of size len bytes.

unsafe fn zeroize_mem(&self, ptr: *mut u8, len: usize)

Zeroize the memory pointed to by ptr and of size len bytes. Shorthand for Self::zeroize_mem_blocks::<0, 0>.

impl Copy for X86_64Sse2Zeroizer