Change the compiler's big integer library to use libTomMath

This now replaces Bill's crappy big int implementation

src/libtommath/mp_dr_setup.c

@@ -0,0 +1,15 @@
+#include "tommath_private.h"
+#ifdef MP_DR_SETUP_C
+/* LibTomMath, multiple-precision integer library -- Tom St Denis */
+/* SPDX-License-Identifier: Unlicense */
+
+/* determines the setup value */
+void mp_dr_setup(const mp_int *a, mp_digit *d)
+{
+   /* the casts are required if MP_DIGIT_BIT is one less than
+    * the number of bits in a mp_digit [e.g. MP_DIGIT_BIT==31]
+    */
+   *d = (mp_digit)(((mp_word)1 << (mp_word)MP_DIGIT_BIT) - (mp_word)a->dp[0]);
+}
+
+#endif