Skip to content

Small performance improvement of mbedtls_mpi_div_mpi() #308

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Nov 19, 2019
Merged

Small performance improvement of mbedtls_mpi_div_mpi() #308

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
35d6d46
Small performance improvement of mbedtls_mpi_div_mpi():
krizhanovsky Oct 31, 2019
d19a193
Fix code review comments:
krizhanovsky Nov 1, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
26 changes: 18 additions & 8 deletions library/bignum.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1632,14 +1632,25 @@ int mbedtls_mpi_div_mpi( mbedtls_mpi *Q, mbedtls_mpi *R, const mbedtls_mpi *A,
int ret;
size_t i, n, t, k;
mbedtls_mpi X, Y, Z, T1, T2;
mbedtls_mpi_uint TP2[3];
MPI_VALIDATE_RET( A != NULL );
MPI_VALIDATE_RET( B != NULL );

if( mbedtls_mpi_cmp_int( B, 0 ) == 0 )
return( MBEDTLS_ERR_MPI_DIVISION_BY_ZERO );

mbedtls_mpi_init( &X ); mbedtls_mpi_init( &Y ); mbedtls_mpi_init( &Z );
mbedtls_mpi_init( &T1 ); mbedtls_mpi_init( &T2 );
mbedtls_mpi_init( &T1 );
/*
* Avoid dynamic memory allocations for constant-size T2.
*
* T2 is used for comparison only and the 3 limbs are assigned explicitly,
* so nobody increase the size of the MPI and we're safe to use an on-stack
* buffer.
*/
T2.s = 1;
T2.n = sizeof( TP2 ) / sizeof( *TP2 );
T2.p = TP2;

if( mbedtls_mpi_cmp_abs( A, B ) < 0 )
{
Expand All @@ -1655,7 +1666,6 @@ int mbedtls_mpi_div_mpi( mbedtls_mpi *Q, mbedtls_mpi *R, const mbedtls_mpi *A,
MBEDTLS_MPI_CHK( mbedtls_mpi_grow( &Z, A->n + 2 ) );
MBEDTLS_MPI_CHK( mbedtls_mpi_lset( &Z, 0 ) );
MBEDTLS_MPI_CHK( mbedtls_mpi_grow( &T1, 2 ) );
MBEDTLS_MPI_CHK( mbedtls_mpi_grow( &T2, 3 ) );

k = mbedtls_mpi_bitlen( &Y ) % biL;
if( k < biL - 1 )
Expand Down Expand Up @@ -1687,6 +1697,10 @@ int mbedtls_mpi_div_mpi( mbedtls_mpi *Q, mbedtls_mpi *R, const mbedtls_mpi *A,
Y.p[t], NULL);
}

T2.p[0] = ( i < 2 ) ? 0 : X.p[i - 2];
T2.p[1] = ( i < 1 ) ? 0 : X.p[i - 1];
T2.p[2] = X.p[i];

Z.p[i - t - 1]++;
do
{
Expand All @@ -1696,11 +1710,6 @@ int mbedtls_mpi_div_mpi( mbedtls_mpi *Q, mbedtls_mpi *R, const mbedtls_mpi *A,
T1.p[0] = ( t < 1 ) ? 0 : Y.p[t - 1];
T1.p[1] = Y.p[t];
MBEDTLS_MPI_CHK( mbedtls_mpi_mul_int( &T1, &T1, Z.p[i - t - 1] ) );

MBEDTLS_MPI_CHK( mbedtls_mpi_lset( &T2, 0 ) );
T2.p[0] = ( i < 2 ) ? 0 : X.p[i - 2];
T2.p[1] = ( i < 1 ) ? 0 : X.p[i - 1];
T2.p[2] = X.p[i];
}
while( mbedtls_mpi_cmp_mpi( &T1, &T2 ) > 0 );

Expand Down Expand Up @@ -1736,7 +1745,8 @@ int mbedtls_mpi_div_mpi( mbedtls_mpi *Q, mbedtls_mpi *R, const mbedtls_mpi *A,
cleanup:

mbedtls_mpi_free( &X ); mbedtls_mpi_free( &Y ); mbedtls_mpi_free( &Z );
mbedtls_mpi_free( &T1 ); mbedtls_mpi_free( &T2 );
mbedtls_mpi_free( &T1 );
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Blocker: __tp2 contains potentially sensitive data. It needs to be wiped with mbedtls_platform_zeroize.

mbedtls_platform_zeroize( TP2, sizeof( TP2 ) );

return( ret );
}
Expand Down