mbedtls hardware bignum: Use memcpy instead of REG_WRITE/REG_READ in a loop

Removes memory barriers for better performance, thanks Ivan for pointing
this out.

Manually unrolling the loop further seemed like diminishing returns.
pull/124/head
Angus Gratton 2016-11-21 18:08:22 +11:00
rodzic 1cc0b3000b
commit 1d47755588
1 zmienionych plików z 17 dodań i 9 usunięć

Wyświetl plik

@ -121,12 +121,16 @@ static inline size_t bits_to_hardware_words(size_t num_bits)
*/
static inline void mpi_to_mem_block(uint32_t mem_base, const mbedtls_mpi *mpi, size_t num_words)
{
for(size_t i = 0; i < mpi->n && i < num_words; i++) {
REG_WRITE(mem_base + i * 4, mpi->p[i]);
}
for(size_t i = mpi->n; i < num_words; i++) {
REG_WRITE(mem_base + i * 4, 0);
}
uint32_t *pbase = (uint32_t *)mem_base;
uint32_t copy_words = num_words < mpi->n ? num_words : mpi->n;
/* Copy MPI data to memory block registers */
memcpy(pbase, mpi->p, copy_words * 4);
/* Zero any remaining memory block data */
bzero(pbase + copy_words, (num_words - copy_words) * 4);
/* Note: not executing memw here, can do it before we start a bignum operation */
}
/* Read mbedTLS MPI bignum back from hardware memory block.
@ -141,15 +145,16 @@ static inline int mem_block_to_mpi(mbedtls_mpi *x, uint32_t mem_base, int num_wo
MBEDTLS_MPI_CHK( mbedtls_mpi_grow(x, num_words) );
for(int i = 0; i < num_words; i++) {
x->p[i] = REG_READ(mem_base + i * 4);
}
/* Copy data from memory block registers */
memcpy(x->p, (uint32_t *)mem_base, num_words * 4);
/* Zero any remaining limbs in the bignum, if the buffer is bigger
than num_words */
for(size_t i = num_words; i < x->n; i++) {
x->p[i] = 0;
}
asm volatile ("memw");
cleanup:
return ret;
}
@ -218,6 +223,9 @@ static inline void execute_op(uint32_t op_reg)
/* Clear interrupt status */
REG_WRITE(RSA_INTERRUPT_REG, 1);
/* Note: above REG_WRITE includes a memw, so we know any writes
to the memory blocks are also complete. */
REG_WRITE(op_reg, 1);
#ifdef CONFIG_MBEDTLS_MPI_USE_INTERRUPT