Make mbedtls_mpi_lt_mpi_ct more portable
The code relied on the assumptions that CHAR_BIT is 8 and that unsigned
does not have padding bits.
In the Bignum module we already assume that the sign of an MPI is either
-1 or 1. Using this, we eliminate the above mentioned dependency.
diff --git a/library/bignum.c b/library/bignum.c
index d310adb..9070486 100644
--- a/library/bignum.c
+++ b/library/bignum.c
@@ -1200,12 +1200,11 @@
return MBEDTLS_ERR_MPI_BAD_INPUT_DATA;
/*
- * Get sign bits of the signs.
+ * Set sign_N to 1 if N >= 0, 0 if N < 0.
+ * We know that N->s == 1 if N >= 0 and N->s == -1 if N < 0.
*/
- sign_X = X->s;
- sign_X = sign_X >> ( sizeof( unsigned ) * 8 - 1 );
- sign_Y = Y->s;
- sign_Y = sign_Y >> ( sizeof( unsigned ) * 8 - 1 );
+ sign_X = ( X->s & 2 ) >> 1;
+ sign_Y = ( Y->s & 2 ) >> 1;
/*
* If the signs are different, then the positive operand is the bigger.