- Introduction
- Library Architecture
- Tpm BigNum Library
- Tpm Crypto Libraries
- Tpm Math Library
- Tpm Platform API
This document describes the TpmBigNum Math library, which is the Reference
Code implementation of the MATH_LIB interface. The detailed MATH_LIB interface
is described in TPM Math Library. This document
describes the internal implementation of TpmBigNum.
In particular, TpmBigNum consumes and requires a helper asymmetric math
library to provide complex big-number math operations. The reference code
currently supports wolfssl and OpenSSL libraries as sub-providers of large
number crypto operations.
TpmBigNum uses the bigNum, and bigPoint types to hold numeric information.
TpmBigNum respects the BIG_ENDIAN_TPM macro to reflect how data is stored in
memory. TpmBigNum provides all the stubs required by the Tpm Math Library
interface, and some of the (constant and linear time) bigNum math operations.
More complex operations such as exponentiation, division, multiplication, and
modular operations are provided by the sub-library. All sub-libraries must use
the same bigNum format for the underlying memory and satisfy the
BnSupport_interface.h interface. The historical strategy is to provide wrapper
functions that convert in and out of the library-native format into the bigNum
library format automatically. If a library wishes to avoid the overhead of
these conversions then they can target the entire MATH_LIB interface instead.
Similar to the Core library's use of MATH_LIB, TpmBigNum also uses a
preprocessor macro to indicate the sub-library it will use.
| Build Flag | Header Name | Description |
|---|---|---|
| BN_MATH_LIB | BnTo[BN_MATH_LIB]Math.h |
Math Interface used to provide bigNum implementations |