PolarSSL is now part of ARM Official announcement and rebranded as Mbed TLS.

Help with ECC and BigNum


Jan 4, 2018 11:08
Antonio

Hi all,

I want to enable Elliptic Cryptography, however I have problem.. my BigNum module uses 16 bit numbers while the ECP modules use 32 bits... when I try to compile I get errors about truncated numbers.. " error: large integer implicitly truncated to unsigned type [-Werror=overflow] ( (mbedtls_mpi_uint) a << 0 ) |" because my "mbedtls_mpi_uint" is 16 bits.

 
Jan 9, 2018 16:05
Ron Eldor

Hi Antonio,
Mbed TLS, does not support 16 bit architecture. If the issue you are encountering is only on the bignum module, I would suggest you set mbedtls_mpi_uint to a 32 bit type. However, I would suspect you will encounter other issues related to 16 bit platforms, in other modules.
Regards,
Mbed TLS Team member
Ron