If you do not understand them, do not not use those functions.
.Pp
.Fn crypto_eddsa_trim_scalar
-reads a 256-bit number represented in Little Endian,
+reads a 256-bit number represented in little-endian,
and outputs the same number modified as follows:
the 3 least significant bits are cleared;
the most significant bit is cleared;
and the second most significant bit is set.
.Pp
.Fn crypto_eddsa_reduce
-reads a 512-bit number represented in Little Endian,
+reads a 512-bit number represented in little-endian,
and outputs the same number reduced modulo the prime order of
Curve25519.
.Pp
.Fn crypto_eddsa_mul_add
-reads three 256-bit numbers represented in Little Endian,
+reads three 256-bit numbers represented in little-endian,
and outputs
.Fa a
×
reduced modulo the prime order of Curve25519.
.Pp
.Fn crypto_eddsa_scalarbase
-reads a 256-bit number represented in Little Endian,
+reads a 256-bit number represented in little-endian,
and outputs the result of the scalar multiplication between that number
and the twisted Edwards base point of Curve25519.
The output uses the same compressed representation as regular EdDSA
coordinate (1 if it is odd, 0 if it is even),
and the 255 other bits represent the
.Em y
-coordinate in Little Endian.
+coordinate in little-endian.
.Pp
.Fn crypto_eddsa_check_equation
reads a signature