krb5 commit: Update edwards25519 fiat code from BoringSSL

ghudson at mit.edu ghudson at mit.edu
Fri Apr 25 00:39:26 EDT 2025


https://github.com/krb5/krb5/commit/2c046a61163a9e969e16fc8644883f8f1d400561
commit 2c046a61163a9e969e16fc8644883f8f1d400561
Author: Greg Hudson <ghudson at mit.edu>
Date:   Wed Apr 16 22:04:58 2025 -0400

    Update edwards25519 fiat code from BoringSSL
    
    Bring the SPAKE edwards25519_fiat.h up to date with BoringSSL commit
    5813c2c10c73d800f1b0d890a7d74ff973abbffc.  This update adds value
    barriers to the conditional move functions, to prevent clang
    optimizations that would make them not constant-time.
    
    Update the path reference comments in edwards25519.c and
    edwards25519_tables.h, but do not update the code as the upstream
    license has changed.

 src/plugins/preauth/spake/edwards25519.c        |    2 +-
 src/plugins/preauth/spake/edwards25519_fiat.h   | 2814 +++++++++++++++--------
 src/plugins/preauth/spake/edwards25519_tables.h |    2 +-
 3 files changed, 1874 insertions(+), 944 deletions(-)

diff --git a/src/plugins/preauth/spake/edwards25519.c b/src/plugins/preauth/spake/edwards25519.c
index 20394115b..77aba1261 100644
--- a/src/plugins/preauth/spake/edwards25519.c
+++ b/src/plugins/preauth/spake/edwards25519.c
@@ -118,7 +118,7 @@ typedef __int128_t int128_t;
 typedef __uint128_t uint128_t;
 #endif
 
-/* From BoringSSL third-party/fiat/internal.h */
+/* From BoringSSL crypto/curve25519/internal.h */
 
 #if defined(BORINGSSL_CURVE25519_64BIT)
 // fe means field element. Here the field is \Z/(2^255-19). An element t,
diff --git a/src/plugins/preauth/spake/edwards25519_fiat.h b/src/plugins/preauth/spake/edwards25519_fiat.h
index 48ade9210..dc4a7ef5f 100644
--- a/src/plugins/preauth/spake/edwards25519_fiat.h
+++ b/src/plugins/preauth/spake/edwards25519_fiat.h
@@ -1,21 +1,62 @@
 #if defined(BORINGSSL_CURVE25519_64BIT)
 
-/* Autogenerated */
+/* Autogenerated: 'src/ExtractionOCaml/unsaturated_solinas' --inline --static --use-value-barrier 25519 64 '(auto)' '2^255 - 19' carry_mul carry_square carry add sub opp selectznz to_bytes from_bytes relax carry_scmul121666 */
 /* curve description: 25519 */
-/* requested operations: carry_mul, carry_square, carry_scmul121666, carry, add, sub, opp, selectznz, to_bytes, from_bytes */
-/* n = 5 (from "5") */
-/* s = 0x8000000000000000000000000000000000000000000000000000000000000000 (from "2^255") */
-/* c = [(1, 19)] (from "1,19") */
 /* machine_wordsize = 64 (from "64") */
+/* requested operations: carry_mul, carry_square, carry, add, sub, opp, selectznz, to_bytes, from_bytes, relax, carry_scmul121666 */
+/* n = 5 (from "(auto)") */
+/* s-c = 2^255 - [(1, 19)] (from "2^255 - 19") */
+/* tight_bounds_multiplier = 1 (from "") */
+/*  */
+/* Computed values: */
+/*   carry_chain = [0, 1, 2, 3, 4, 0, 1] */
+/*   eval z = z[0] + (z[1] << 51) + (z[2] << 102) + (z[3] << 153) + (z[4] << 204) */
+/*   bytes_eval z = z[0] + (z[1] << 8) + (z[2] << 16) + (z[3] << 24) + (z[4] << 32) + (z[5] << 40) + (z[6] << 48) + (z[7] << 56) + (z[8] << 64) + (z[9] << 72) + (z[10] << 80) + (z[11] << 88) + (z[12] << 96) + (z[13] << 104) + (z[14] << 112) + (z[15] << 120) + (z[16] << 128) + (z[17] << 136) + (z[18] << 144) + (z[19] << 152) + (z[20] << 160) + (z[21] << 168) + (z[22] << 176) + (z[23] << 184) + (z[24] << 192) + (z[25] << 200) + (z[26] << 208) + (z[27] << 216) + (z[28] << 224) + (z[29] << 232) + (z[30] << 240) + (z[31] << 248) */
+/*   balance = [0xfffffffffffda, 0xffffffffffffe, 0xffffffffffffe, 0xffffffffffffe, 0xffffffffffffe] */
 
 #include <stdint.h>
 typedef unsigned char fiat_25519_uint1;
 typedef signed char fiat_25519_int1;
-typedef int128_t fiat_25519_int128;
-typedef uint128_t fiat_25519_uint128;
+#if defined(__GNUC__) || defined(__clang__)
+#  define FIAT_25519_FIAT_EXTENSION __extension__
+#  define FIAT_25519_FIAT_INLINE __inline__
+#else
+#  define FIAT_25519_FIAT_EXTENSION
+#  define FIAT_25519_FIAT_INLINE
+#endif
+
+FIAT_25519_FIAT_EXTENSION typedef int128_t fiat_25519_int128;
+FIAT_25519_FIAT_EXTENSION typedef uint128_t fiat_25519_uint128;
+
+/* The type fiat_25519_loose_field_element is a field element with loose bounds. */
+/* Bounds: [[0x0 ~> 0x18000000000000], [0x0 ~> 0x18000000000000], [0x0 ~> 0x18000000000000], [0x0 ~> 0x18000000000000], [0x0 ~> 0x18000000000000]] */
+typedef uint64_t fiat_25519_loose_field_element[5];
+
+/* The type fiat_25519_tight_field_element is a field element with tight bounds. */
+/* Bounds: [[0x0 ~> 0x8000000000000], [0x0 ~> 0x8000000000000], [0x0 ~> 0x8000000000000], [0x0 ~> 0x8000000000000], [0x0 ~> 0x8000000000000]] */
+typedef uint64_t fiat_25519_tight_field_element[5];
+
+#if (-1 & 3) != 3
+#error "This code only works on a two's complement system"
+#endif
+
+#if !defined(FIAT_25519_NO_ASM) && (defined(__GNUC__) || defined(__clang__))
+static __inline__ uint64_t fiat_25519_value_barrier_u64(uint64_t a) {
+  __asm__("" : "+r"(a) : /* no inputs */);
+  return a;
+}
+#else
+#  define fiat_25519_value_barrier_u64(x) (x)
+#endif
 
 
 /*
+ * The function fiat_25519_addcarryx_u51 is an addition with carry.
+ *
+ * Postconditions:
+ *   out1 = (arg1 + arg2 + arg3) mod 2^51
+ *   out2 = ⌊(arg1 + arg2 + arg3) / 2^51⌋
+ *
  * Input Bounds:
  *   arg1: [0x0 ~> 0x1]
  *   arg2: [0x0 ~> 0x7ffffffffffff]
@@ -24,15 +65,24 @@ typedef uint128_t fiat_25519_uint128;
  *   out1: [0x0 ~> 0x7ffffffffffff]
  *   out2: [0x0 ~> 0x1]
  */
-static void fiat_25519_addcarryx_u51(uint64_t* out1, fiat_25519_uint1* out2, fiat_25519_uint1 arg1, uint64_t arg2, uint64_t arg3) {
-  uint64_t x1 = ((arg1 + arg2) + arg3);
-  uint64_t x2 = (x1 & UINT64_C(0x7ffffffffffff));
-  fiat_25519_uint1 x3 = (fiat_25519_uint1)(x1 >> 51);
+static FIAT_25519_FIAT_INLINE void fiat_25519_addcarryx_u51(uint64_t* out1, fiat_25519_uint1* out2, fiat_25519_uint1 arg1, uint64_t arg2, uint64_t arg3) {
+  uint64_t x1;
+  uint64_t x2;
+  fiat_25519_uint1 x3;
+  x1 = ((arg1 + arg2) + arg3);
+  x2 = (x1 & UINT64_C(0x7ffffffffffff));
+  x3 = (fiat_25519_uint1)(x1 >> 51);
   *out1 = x2;
   *out2 = x3;
 }
 
 /*
+ * The function fiat_25519_subborrowx_u51 is a subtraction with borrow.
+ *
+ * Postconditions:
+ *   out1 = (-arg1 + arg2 + -arg3) mod 2^51
+ *   out2 = -⌊(-arg1 + arg2 + -arg3) / 2^51⌋
+ *
  * Input Bounds:
  *   arg1: [0x0 ~> 0x1]
  *   arg2: [0x0 ~> 0x7ffffffffffff]
@@ -41,15 +91,23 @@ static void fiat_25519_addcarryx_u51(uint64_t* out1, fiat_25519_uint1* out2, fia
  *   out1: [0x0 ~> 0x7ffffffffffff]
  *   out2: [0x0 ~> 0x1]
  */
-static void fiat_25519_subborrowx_u51(uint64_t* out1, fiat_25519_uint1* out2, fiat_25519_uint1 arg1, uint64_t arg2, uint64_t arg3) {
-  int64_t x1 = ((int64_t)(arg2 - (int64_t)arg1) - (int64_t)arg3);
-  fiat_25519_int1 x2 = (fiat_25519_int1)(x1 >> 51);
-  uint64_t x3 = (x1 & UINT64_C(0x7ffffffffffff));
+static FIAT_25519_FIAT_INLINE void fiat_25519_subborrowx_u51(uint64_t* out1, fiat_25519_uint1* out2, fiat_25519_uint1 arg1, uint64_t arg2, uint64_t arg3) {
+  int64_t x1;
+  fiat_25519_int1 x2;
+  uint64_t x3;
+  x1 = ((int64_t)(arg2 - (int64_t)arg1) - (int64_t)arg3);
+  x2 = (fiat_25519_int1)(x1 >> 51);
+  x3 = (x1 & UINT64_C(0x7ffffffffffff));
   *out1 = x3;
   *out2 = (fiat_25519_uint1)(0x0 - x2);
 }
 
 /*
+ * The function fiat_25519_cmovznz_u64 is a single-word conditional move.
+ *
+ * Postconditions:
+ *   out1 = (if arg1 = 0 then arg2 else arg3)
+ *
  * Input Bounds:
  *   arg1: [0x0 ~> 0x1]
  *   arg2: [0x0 ~> 0xffffffffffffffff]
@@ -57,73 +115,128 @@ static void fiat_25519_subborrowx_u51(uint64_t* out1, fiat_25519_uint1* out2, fi
  * Output Bounds:
  *   out1: [0x0 ~> 0xffffffffffffffff]
  */
-static void fiat_25519_cmovznz_u64(uint64_t* out1, fiat_25519_uint1 arg1, uint64_t arg2, uint64_t arg3) {
-  fiat_25519_uint1 x1 = (!(!arg1));
-  uint64_t x2 = ((fiat_25519_int1)(0x0 - x1) & UINT64_C(0xffffffffffffffff));
-  uint64_t x3 = ((x2 & arg3) | ((~x2) & arg2));
+static FIAT_25519_FIAT_INLINE void fiat_25519_cmovznz_u64(uint64_t* out1, fiat_25519_uint1 arg1, uint64_t arg2, uint64_t arg3) {
+  fiat_25519_uint1 x1;
+  uint64_t x2;
+  uint64_t x3;
+  x1 = (!(!arg1));
+  x2 = ((fiat_25519_int1)(0x0 - x1) & UINT64_C(0xffffffffffffffff));
+  x3 = ((fiat_25519_value_barrier_u64(x2) & arg3) | (fiat_25519_value_barrier_u64((~x2)) & arg2));
   *out1 = x3;
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664]]
- *   arg2: [[0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc]]
+ * The function fiat_25519_carry_mul multiplies two field elements and reduces the result.
+ *
+ * Postconditions:
+ *   eval out1 mod m = (eval arg1 * eval arg2) mod m
+ *
  */
-static void fiat_25519_carry_mul(uint64_t out1[5], const uint64_t arg1[5], const uint64_t arg2[5]) {
-  fiat_25519_uint128 x1 = ((fiat_25519_uint128)(arg1[4]) * ((arg2[4]) * (uint64_t)UINT8_C(0x13)));
-  fiat_25519_uint128 x2 = ((fiat_25519_uint128)(arg1[4]) * ((arg2[3]) * (uint64_t)UINT8_C(0x13)));
-  fiat_25519_uint128 x3 = ((fiat_25519_uint128)(arg1[4]) * ((arg2[2]) * (uint64_t)UINT8_C(0x13)));
-  fiat_25519_uint128 x4 = ((fiat_25519_uint128)(arg1[4]) * ((arg2[1]) * (uint64_t)UINT8_C(0x13)));
-  fiat_25519_uint128 x5 = ((fiat_25519_uint128)(arg1[3]) * ((arg2[4]) * (uint64_t)UINT8_C(0x13)));
-  fiat_25519_uint128 x6 = ((fiat_25519_uint128)(arg1[3]) * ((arg2[3]) * (uint64_t)UINT8_C(0x13)));
-  fiat_25519_uint128 x7 = ((fiat_25519_uint128)(arg1[3]) * ((arg2[2]) * (uint64_t)UINT8_C(0x13)));
-  fiat_25519_uint128 x8 = ((fiat_25519_uint128)(arg1[2]) * ((arg2[4]) * (uint64_t)UINT8_C(0x13)));
-  fiat_25519_uint128 x9 = ((fiat_25519_uint128)(arg1[2]) * ((arg2[3]) * (uint64_t)UINT8_C(0x13)));
-  fiat_25519_uint128 x10 = ((fiat_25519_uint128)(arg1[1]) * ((arg2[4]) * (uint64_t)UINT8_C(0x13)));
-  fiat_25519_uint128 x11 = ((fiat_25519_uint128)(arg1[4]) * (arg2[0]));
-  fiat_25519_uint128 x12 = ((fiat_25519_uint128)(arg1[3]) * (arg2[1]));
-  fiat_25519_uint128 x13 = ((fiat_25519_uint128)(arg1[3]) * (arg2[0]));
-  fiat_25519_uint128 x14 = ((fiat_25519_uint128)(arg1[2]) * (arg2[2]));
-  fiat_25519_uint128 x15 = ((fiat_25519_uint128)(arg1[2]) * (arg2[1]));
-  fiat_25519_uint128 x16 = ((fiat_25519_uint128)(arg1[2]) * (arg2[0]));
-  fiat_25519_uint128 x17 = ((fiat_25519_uint128)(arg1[1]) * (arg2[3]));
-  fiat_25519_uint128 x18 = ((fiat_25519_uint128)(arg1[1]) * (arg2[2]));
-  fiat_25519_uint128 x19 = ((fiat_25519_uint128)(arg1[1]) * (arg2[1]));
-  fiat_25519_uint128 x20 = ((fiat_25519_uint128)(arg1[1]) * (arg2[0]));
-  fiat_25519_uint128 x21 = ((fiat_25519_uint128)(arg1[0]) * (arg2[4]));
-  fiat_25519_uint128 x22 = ((fiat_25519_uint128)(arg1[0]) * (arg2[3]));
-  fiat_25519_uint128 x23 = ((fiat_25519_uint128)(arg1[0]) * (arg2[2]));
-  fiat_25519_uint128 x24 = ((fiat_25519_uint128)(arg1[0]) * (arg2[1]));
-  fiat_25519_uint128 x25 = ((fiat_25519_uint128)(arg1[0]) * (arg2[0]));
-  fiat_25519_uint128 x26 = (x25 + (x10 + (x9 + (x7 + x4))));
-  uint64_t x27 = (uint64_t)(x26 >> 51);
-  uint64_t x28 = (uint64_t)(x26 & UINT64_C(0x7ffffffffffff));
-  fiat_25519_uint128 x29 = (x21 + (x17 + (x14 + (x12 + x11))));
-  fiat_25519_uint128 x30 = (x22 + (x18 + (x15 + (x13 + x1))));
-  fiat_25519_uint128 x31 = (x23 + (x19 + (x16 + (x5 + x2))));
-  fiat_25519_uint128 x32 = (x24 + (x20 + (x8 + (x6 + x3))));
-  fiat_25519_uint128 x33 = (x27 + x32);
-  uint64_t x34 = (uint64_t)(x33 >> 51);
-  uint64_t x35 = (uint64_t)(x33 & UINT64_C(0x7ffffffffffff));
-  fiat_25519_uint128 x36 = (x34 + x31);
-  uint64_t x37 = (uint64_t)(x36 >> 51);
-  uint64_t x38 = (uint64_t)(x36 & UINT64_C(0x7ffffffffffff));
-  fiat_25519_uint128 x39 = (x37 + x30);
-  uint64_t x40 = (uint64_t)(x39 >> 51);
-  uint64_t x41 = (uint64_t)(x39 & UINT64_C(0x7ffffffffffff));
-  fiat_25519_uint128 x42 = (x40 + x29);
-  uint64_t x43 = (uint64_t)(x42 >> 51);
-  uint64_t x44 = (uint64_t)(x42 & UINT64_C(0x7ffffffffffff));
-  uint64_t x45 = (x43 * (uint64_t)UINT8_C(0x13));
-  uint64_t x46 = (x28 + x45);
-  uint64_t x47 = (x46 >> 51);
-  uint64_t x48 = (x46 & UINT64_C(0x7ffffffffffff));
-  uint64_t x49 = (x47 + x35);
-  uint64_t x50 = (x49 >> 51);
-  uint64_t x51 = (x49 & UINT64_C(0x7ffffffffffff));
-  uint64_t x52 = (x50 + x38);
+static FIAT_25519_FIAT_INLINE void fiat_25519_carry_mul(fiat_25519_tight_field_element out1, const fiat_25519_loose_field_element arg1, const fiat_25519_loose_field_element arg2) {
+  fiat_25519_uint128 x1;
+  fiat_25519_uint128 x2;
+  fiat_25519_uint128 x3;
+  fiat_25519_uint128 x4;
+  fiat_25519_uint128 x5;
+  fiat_25519_uint128 x6;
+  fiat_25519_uint128 x7;
+  fiat_25519_uint128 x8;
+  fiat_25519_uint128 x9;
+  fiat_25519_uint128 x10;
+  fiat_25519_uint128 x11;
+  fiat_25519_uint128 x12;
+  fiat_25519_uint128 x13;
+  fiat_25519_uint128 x14;
+  fiat_25519_uint128 x15;
+  fiat_25519_uint128 x16;
+  fiat_25519_uint128 x17;
+  fiat_25519_uint128 x18;
+  fiat_25519_uint128 x19;
+  fiat_25519_uint128 x20;
+  fiat_25519_uint128 x21;
+  fiat_25519_uint128 x22;
+  fiat_25519_uint128 x23;
+  fiat_25519_uint128 x24;
+  fiat_25519_uint128 x25;
+  fiat_25519_uint128 x26;
+  uint64_t x27;
+  uint64_t x28;
+  fiat_25519_uint128 x29;
+  fiat_25519_uint128 x30;
+  fiat_25519_uint128 x31;
+  fiat_25519_uint128 x32;
+  fiat_25519_uint128 x33;
+  uint64_t x34;
+  uint64_t x35;
+  fiat_25519_uint128 x36;
+  uint64_t x37;
+  uint64_t x38;
+  fiat_25519_uint128 x39;
+  uint64_t x40;
+  uint64_t x41;
+  fiat_25519_uint128 x42;
+  uint64_t x43;
+  uint64_t x44;
+  uint64_t x45;
+  uint64_t x46;
+  uint64_t x47;
+  uint64_t x48;
+  uint64_t x49;
+  fiat_25519_uint1 x50;
+  uint64_t x51;
+  uint64_t x52;
+  x1 = ((fiat_25519_uint128)(arg1[4]) * ((arg2[4]) * UINT8_C(0x13)));
+  x2 = ((fiat_25519_uint128)(arg1[4]) * ((arg2[3]) * UINT8_C(0x13)));
+  x3 = ((fiat_25519_uint128)(arg1[4]) * ((arg2[2]) * UINT8_C(0x13)));
+  x4 = ((fiat_25519_uint128)(arg1[4]) * ((arg2[1]) * UINT8_C(0x13)));
+  x5 = ((fiat_25519_uint128)(arg1[3]) * ((arg2[4]) * UINT8_C(0x13)));
+  x6 = ((fiat_25519_uint128)(arg1[3]) * ((arg2[3]) * UINT8_C(0x13)));
+  x7 = ((fiat_25519_uint128)(arg1[3]) * ((arg2[2]) * UINT8_C(0x13)));
+  x8 = ((fiat_25519_uint128)(arg1[2]) * ((arg2[4]) * UINT8_C(0x13)));
+  x9 = ((fiat_25519_uint128)(arg1[2]) * ((arg2[3]) * UINT8_C(0x13)));
+  x10 = ((fiat_25519_uint128)(arg1[1]) * ((arg2[4]) * UINT8_C(0x13)));
+  x11 = ((fiat_25519_uint128)(arg1[4]) * (arg2[0]));
+  x12 = ((fiat_25519_uint128)(arg1[3]) * (arg2[1]));
+  x13 = ((fiat_25519_uint128)(arg1[3]) * (arg2[0]));
+  x14 = ((fiat_25519_uint128)(arg1[2]) * (arg2[2]));
+  x15 = ((fiat_25519_uint128)(arg1[2]) * (arg2[1]));
+  x16 = ((fiat_25519_uint128)(arg1[2]) * (arg2[0]));
+  x17 = ((fiat_25519_uint128)(arg1[1]) * (arg2[3]));
+  x18 = ((fiat_25519_uint128)(arg1[1]) * (arg2[2]));
+  x19 = ((fiat_25519_uint128)(arg1[1]) * (arg2[1]));
+  x20 = ((fiat_25519_uint128)(arg1[1]) * (arg2[0]));
+  x21 = ((fiat_25519_uint128)(arg1[0]) * (arg2[4]));
+  x22 = ((fiat_25519_uint128)(arg1[0]) * (arg2[3]));
+  x23 = ((fiat_25519_uint128)(arg1[0]) * (arg2[2]));
+  x24 = ((fiat_25519_uint128)(arg1[0]) * (arg2[1]));
+  x25 = ((fiat_25519_uint128)(arg1[0]) * (arg2[0]));
+  x26 = (x25 + (x10 + (x9 + (x7 + x4))));
+  x27 = (uint64_t)(x26 >> 51);
+  x28 = (uint64_t)(x26 & UINT64_C(0x7ffffffffffff));
+  x29 = (x21 + (x17 + (x14 + (x12 + x11))));
+  x30 = (x22 + (x18 + (x15 + (x13 + x1))));
+  x31 = (x23 + (x19 + (x16 + (x5 + x2))));
+  x32 = (x24 + (x20 + (x8 + (x6 + x3))));
+  x33 = (x27 + x32);
+  x34 = (uint64_t)(x33 >> 51);
+  x35 = (uint64_t)(x33 & UINT64_C(0x7ffffffffffff));
+  x36 = (x34 + x31);
+  x37 = (uint64_t)(x36 >> 51);
+  x38 = (uint64_t)(x36 & UINT64_C(0x7ffffffffffff));
+  x39 = (x37 + x30);
+  x40 = (uint64_t)(x39 >> 51);
+  x41 = (uint64_t)(x39 & UINT64_C(0x7ffffffffffff));
+  x42 = (x40 + x29);
+  x43 = (uint64_t)(x42 >> 51);
+  x44 = (uint64_t)(x42 & UINT64_C(0x7ffffffffffff));
+  x45 = (x43 * UINT8_C(0x13));
+  x46 = (x28 + x45);
+  x47 = (x46 >> 51);
+  x48 = (x46 & UINT64_C(0x7ffffffffffff));
+  x49 = (x47 + x35);
+  x50 = (fiat_25519_uint1)(x49 >> 51);
+  x51 = (x49 & UINT64_C(0x7ffffffffffff));
+  x52 = (x50 + x38);
   out1[0] = x48;
   out1[1] = x51;
   out1[2] = x52;
@@ -132,62 +245,113 @@ static void fiat_25519_carry_mul(uint64_t out1[5], const uint64_t arg1[5], const
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc]]
+ * The function fiat_25519_carry_square squares a field element and reduces the result.
+ *
+ * Postconditions:
+ *   eval out1 mod m = (eval arg1 * eval arg1) mod m
+ *
  */
-static void fiat_25519_carry_square(uint64_t out1[5], const uint64_t arg1[5]) {
-  uint64_t x1 = ((arg1[4]) * (uint64_t)UINT8_C(0x13));
-  uint64_t x2 = (x1 * (uint64_t)0x2);
-  uint64_t x3 = ((arg1[4]) * (uint64_t)0x2);
-  uint64_t x4 = ((arg1[3]) * (uint64_t)UINT8_C(0x13));
-  uint64_t x5 = (x4 * (uint64_t)0x2);
-  uint64_t x6 = ((arg1[3]) * (uint64_t)0x2);
-  uint64_t x7 = ((arg1[2]) * (uint64_t)0x2);
-  uint64_t x8 = ((arg1[1]) * (uint64_t)0x2);
-  fiat_25519_uint128 x9 = ((fiat_25519_uint128)(arg1[4]) * x1);
-  fiat_25519_uint128 x10 = ((fiat_25519_uint128)(arg1[3]) * x2);
-  fiat_25519_uint128 x11 = ((fiat_25519_uint128)(arg1[3]) * x4);
-  fiat_25519_uint128 x12 = ((fiat_25519_uint128)(arg1[2]) * x2);
-  fiat_25519_uint128 x13 = ((fiat_25519_uint128)(arg1[2]) * x5);
-  fiat_25519_uint128 x14 = ((fiat_25519_uint128)(arg1[2]) * (arg1[2]));
-  fiat_25519_uint128 x15 = ((fiat_25519_uint128)(arg1[1]) * x2);
-  fiat_25519_uint128 x16 = ((fiat_25519_uint128)(arg1[1]) * x6);
-  fiat_25519_uint128 x17 = ((fiat_25519_uint128)(arg1[1]) * x7);
-  fiat_25519_uint128 x18 = ((fiat_25519_uint128)(arg1[1]) * (arg1[1]));
-  fiat_25519_uint128 x19 = ((fiat_25519_uint128)(arg1[0]) * x3);
-  fiat_25519_uint128 x20 = ((fiat_25519_uint128)(arg1[0]) * x6);
-  fiat_25519_uint128 x21 = ((fiat_25519_uint128)(arg1[0]) * x7);
-  fiat_25519_uint128 x22 = ((fiat_25519_uint128)(arg1[0]) * x8);
-  fiat_25519_uint128 x23 = ((fiat_25519_uint128)(arg1[0]) * (arg1[0]));
-  fiat_25519_uint128 x24 = (x23 + (x15 + x13));
-  uint64_t x25 = (uint64_t)(x24 >> 51);
-  uint64_t x26 = (uint64_t)(x24 & UINT64_C(0x7ffffffffffff));
-  fiat_25519_uint128 x27 = (x19 + (x16 + x14));
-  fiat_25519_uint128 x28 = (x20 + (x17 + x9));
-  fiat_25519_uint128 x29 = (x21 + (x18 + x10));
-  fiat_25519_uint128 x30 = (x22 + (x12 + x11));
-  fiat_25519_uint128 x31 = (x25 + x30);
-  uint64_t x32 = (uint64_t)(x31 >> 51);
-  uint64_t x33 = (uint64_t)(x31 & UINT64_C(0x7ffffffffffff));
-  fiat_25519_uint128 x34 = (x32 + x29);
-  uint64_t x35 = (uint64_t)(x34 >> 51);
-  uint64_t x36 = (uint64_t)(x34 & UINT64_C(0x7ffffffffffff));
-  fiat_25519_uint128 x37 = (x35 + x28);
-  uint64_t x38 = (uint64_t)(x37 >> 51);
-  uint64_t x39 = (uint64_t)(x37 & UINT64_C(0x7ffffffffffff));
-  fiat_25519_uint128 x40 = (x38 + x27);
-  uint64_t x41 = (uint64_t)(x40 >> 51);
-  uint64_t x42 = (uint64_t)(x40 & UINT64_C(0x7ffffffffffff));
-  uint64_t x43 = (x41 * (uint64_t)UINT8_C(0x13));
-  uint64_t x44 = (x26 + x43);
-  uint64_t x45 = (x44 >> 51);
-  uint64_t x46 = (x44 & UINT64_C(0x7ffffffffffff));
-  uint64_t x47 = (x45 + x33);
-  uint64_t x48 = (x47 >> 51);
-  uint64_t x49 = (x47 & UINT64_C(0x7ffffffffffff));
-  uint64_t x50 = (x48 + x36);
+static FIAT_25519_FIAT_INLINE void fiat_25519_carry_square(fiat_25519_tight_field_element out1, const fiat_25519_loose_field_element arg1) {
+  uint64_t x1;
+  uint64_t x2;
+  uint64_t x3;
+  uint64_t x4;
+  uint64_t x5;
+  uint64_t x6;
+  uint64_t x7;
+  uint64_t x8;
+  fiat_25519_uint128 x9;
+  fiat_25519_uint128 x10;
+  fiat_25519_uint128 x11;
+  fiat_25519_uint128 x12;
+  fiat_25519_uint128 x13;
+  fiat_25519_uint128 x14;
+  fiat_25519_uint128 x15;
+  fiat_25519_uint128 x16;
+  fiat_25519_uint128 x17;
+  fiat_25519_uint128 x18;
+  fiat_25519_uint128 x19;
+  fiat_25519_uint128 x20;
+  fiat_25519_uint128 x21;
+  fiat_25519_uint128 x22;
+  fiat_25519_uint128 x23;
+  fiat_25519_uint128 x24;
+  uint64_t x25;
+  uint64_t x26;
+  fiat_25519_uint128 x27;
+  fiat_25519_uint128 x28;
+  fiat_25519_uint128 x29;
+  fiat_25519_uint128 x30;
+  fiat_25519_uint128 x31;
+  uint64_t x32;
+  uint64_t x33;
+  fiat_25519_uint128 x34;
+  uint64_t x35;
+  uint64_t x36;
+  fiat_25519_uint128 x37;
+  uint64_t x38;
+  uint64_t x39;
+  fiat_25519_uint128 x40;
+  uint64_t x41;
+  uint64_t x42;
+  uint64_t x43;
+  uint64_t x44;
+  uint64_t x45;
+  uint64_t x46;
+  uint64_t x47;
+  fiat_25519_uint1 x48;
+  uint64_t x49;
+  uint64_t x50;
+  x1 = ((arg1[4]) * UINT8_C(0x13));
+  x2 = (x1 * 0x2);
+  x3 = ((arg1[4]) * 0x2);
+  x4 = ((arg1[3]) * UINT8_C(0x13));
+  x5 = (x4 * 0x2);
+  x6 = ((arg1[3]) * 0x2);
+  x7 = ((arg1[2]) * 0x2);
+  x8 = ((arg1[1]) * 0x2);
+  x9 = ((fiat_25519_uint128)(arg1[4]) * x1);
+  x10 = ((fiat_25519_uint128)(arg1[3]) * x2);
+  x11 = ((fiat_25519_uint128)(arg1[3]) * x4);
+  x12 = ((fiat_25519_uint128)(arg1[2]) * x2);
+  x13 = ((fiat_25519_uint128)(arg1[2]) * x5);
+  x14 = ((fiat_25519_uint128)(arg1[2]) * (arg1[2]));
+  x15 = ((fiat_25519_uint128)(arg1[1]) * x2);
+  x16 = ((fiat_25519_uint128)(arg1[1]) * x6);
+  x17 = ((fiat_25519_uint128)(arg1[1]) * x7);
+  x18 = ((fiat_25519_uint128)(arg1[1]) * (arg1[1]));
+  x19 = ((fiat_25519_uint128)(arg1[0]) * x3);
+  x20 = ((fiat_25519_uint128)(arg1[0]) * x6);
+  x21 = ((fiat_25519_uint128)(arg1[0]) * x7);
+  x22 = ((fiat_25519_uint128)(arg1[0]) * x8);
+  x23 = ((fiat_25519_uint128)(arg1[0]) * (arg1[0]));
+  x24 = (x23 + (x15 + x13));
+  x25 = (uint64_t)(x24 >> 51);
+  x26 = (uint64_t)(x24 & UINT64_C(0x7ffffffffffff));
+  x27 = (x19 + (x16 + x14));
+  x28 = (x20 + (x17 + x9));
+  x29 = (x21 + (x18 + x10));
+  x30 = (x22 + (x12 + x11));
+  x31 = (x25 + x30);
+  x32 = (uint64_t)(x31 >> 51);
+  x33 = (uint64_t)(x31 & UINT64_C(0x7ffffffffffff));
+  x34 = (x32 + x29);
+  x35 = (uint64_t)(x34 >> 51);
+  x36 = (uint64_t)(x34 & UINT64_C(0x7ffffffffffff));
+  x37 = (x35 + x28);
+  x38 = (uint64_t)(x37 >> 51);
+  x39 = (uint64_t)(x37 & UINT64_C(0x7ffffffffffff));
+  x40 = (x38 + x27);
+  x41 = (uint64_t)(x40 >> 51);
+  x42 = (uint64_t)(x40 & UINT64_C(0x7ffffffffffff));
+  x43 = (x41 * UINT8_C(0x13));
+  x44 = (x26 + x43);
+  x45 = (x44 >> 51);
+  x46 = (x44 & UINT64_C(0x7ffffffffffff));
+  x47 = (x45 + x33);
+  x48 = (fiat_25519_uint1)(x47 >> 51);
+  x49 = (x47 & UINT64_C(0x7ffffffffffff));
+  x50 = (x48 + x36);
   out1[0] = x46;
   out1[1] = x49;
   out1[2] = x50;
@@ -196,24 +360,37 @@ static void fiat_25519_carry_square(uint64_t out1[5], const uint64_t arg1[5]) {
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc]]
+ * The function fiat_25519_carry reduces a field element.
+ *
+ * Postconditions:
+ *   eval out1 mod m = eval arg1 mod m
+ *
  */
-static void fiat_25519_carry(uint64_t out1[5], const uint64_t arg1[5]) {
-  uint64_t x1 = (arg1[0]);
-  uint64_t x2 = ((x1 >> 51) + (arg1[1]));
-  uint64_t x3 = ((x2 >> 51) + (arg1[2]));
-  uint64_t x4 = ((x3 >> 51) + (arg1[3]));
-  uint64_t x5 = ((x4 >> 51) + (arg1[4]));
-  uint64_t x6 = ((x1 & UINT64_C(0x7ffffffffffff)) + ((x5 >> 51) * (uint64_t)UINT8_C(0x13)));
-  uint64_t x7 = ((x6 >> 51) + (x2 & UINT64_C(0x7ffffffffffff)));
-  uint64_t x8 = (x6 & UINT64_C(0x7ffffffffffff));
-  uint64_t x9 = (x7 & UINT64_C(0x7ffffffffffff));
-  uint64_t x10 = ((x7 >> 51) + (x3 & UINT64_C(0x7ffffffffffff)));
-  uint64_t x11 = (x4 & UINT64_C(0x7ffffffffffff));
-  uint64_t x12 = (x5 & UINT64_C(0x7ffffffffffff));
+static FIAT_25519_FIAT_INLINE void fiat_25519_carry(fiat_25519_tight_field_element out1, const fiat_25519_loose_field_element arg1) {
+  uint64_t x1;
+  uint64_t x2;
+  uint64_t x3;
+  uint64_t x4;
+  uint64_t x5;
+  uint64_t x6;
+  uint64_t x7;
+  uint64_t x8;
+  uint64_t x9;
+  uint64_t x10;
+  uint64_t x11;
+  uint64_t x12;
+  x1 = (arg1[0]);
+  x2 = ((x1 >> 51) + (arg1[1]));
+  x3 = ((x2 >> 51) + (arg1[2]));
+  x4 = ((x3 >> 51) + (arg1[3]));
+  x5 = ((x4 >> 51) + (arg1[4]));
+  x6 = ((x1 & UINT64_C(0x7ffffffffffff)) + ((x5 >> 51) * UINT8_C(0x13)));
+  x7 = ((fiat_25519_uint1)(x6 >> 51) + (x2 & UINT64_C(0x7ffffffffffff)));
+  x8 = (x6 & UINT64_C(0x7ffffffffffff));
+  x9 = (x7 & UINT64_C(0x7ffffffffffff));
+  x10 = ((fiat_25519_uint1)(x7 >> 51) + (x3 & UINT64_C(0x7ffffffffffff)));
+  x11 = (x4 & UINT64_C(0x7ffffffffffff));
+  x12 = (x5 & UINT64_C(0x7ffffffffffff));
   out1[0] = x8;
   out1[1] = x9;
   out1[2] = x10;
@@ -222,18 +399,23 @@ static void fiat_25519_carry(uint64_t out1[5], const uint64_t arg1[5]) {
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc]]
- *   arg2: [[0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664]]
+ * The function fiat_25519_add adds two field elements.
+ *
+ * Postconditions:
+ *   eval out1 mod m = (eval arg1 + eval arg2) mod m
+ *
  */
-static void fiat_25519_add(uint64_t out1[5], const uint64_t arg1[5], const uint64_t arg2[5]) {
-  uint64_t x1 = ((arg1[0]) + (arg2[0]));
-  uint64_t x2 = ((arg1[1]) + (arg2[1]));
-  uint64_t x3 = ((arg1[2]) + (arg2[2]));
-  uint64_t x4 = ((arg1[3]) + (arg2[3]));
-  uint64_t x5 = ((arg1[4]) + (arg2[4]));
+static FIAT_25519_FIAT_INLINE void fiat_25519_add(fiat_25519_loose_field_element out1, const fiat_25519_tight_field_element arg1, const fiat_25519_tight_field_element arg2) {
+  uint64_t x1;
+  uint64_t x2;
+  uint64_t x3;
+  uint64_t x4;
+  uint64_t x5;
+  x1 = ((arg1[0]) + (arg2[0]));
+  x2 = ((arg1[1]) + (arg2[1]));
+  x3 = ((arg1[2]) + (arg2[2]));
+  x4 = ((arg1[3]) + (arg2[3]));
+  x5 = ((arg1[4]) + (arg2[4]));
   out1[0] = x1;
   out1[1] = x2;
   out1[2] = x3;
@@ -242,18 +424,23 @@ static void fiat_25519_add(uint64_t out1[5], const uint64_t arg1[5], const uint6
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc]]
- *   arg2: [[0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664]]
+ * The function fiat_25519_sub subtracts two field elements.
+ *
+ * Postconditions:
+ *   eval out1 mod m = (eval arg1 - eval arg2) mod m
+ *
  */
-static void fiat_25519_sub(uint64_t out1[5], const uint64_t arg1[5], const uint64_t arg2[5]) {
-  uint64_t x1 = ((UINT64_C(0xfffffffffffda) + (arg1[0])) - (arg2[0]));
-  uint64_t x2 = ((UINT64_C(0xffffffffffffe) + (arg1[1])) - (arg2[1]));
-  uint64_t x3 = ((UINT64_C(0xffffffffffffe) + (arg1[2])) - (arg2[2]));
-  uint64_t x4 = ((UINT64_C(0xffffffffffffe) + (arg1[3])) - (arg2[3]));
-  uint64_t x5 = ((UINT64_C(0xffffffffffffe) + (arg1[4])) - (arg2[4]));
+static FIAT_25519_FIAT_INLINE void fiat_25519_sub(fiat_25519_loose_field_element out1, const fiat_25519_tight_field_element arg1, const fiat_25519_tight_field_element arg2) {
+  uint64_t x1;
+  uint64_t x2;
+  uint64_t x3;
+  uint64_t x4;
+  uint64_t x5;
+  x1 = ((UINT64_C(0xfffffffffffda) + (arg1[0])) - (arg2[0]));
+  x2 = ((UINT64_C(0xffffffffffffe) + (arg1[1])) - (arg2[1]));
+  x3 = ((UINT64_C(0xffffffffffffe) + (arg1[2])) - (arg2[2]));
+  x4 = ((UINT64_C(0xffffffffffffe) + (arg1[3])) - (arg2[3]));
+  x5 = ((UINT64_C(0xffffffffffffe) + (arg1[4])) - (arg2[4]));
   out1[0] = x1;
   out1[1] = x2;
   out1[2] = x3;
@@ -262,17 +449,23 @@ static void fiat_25519_sub(uint64_t out1[5], const uint64_t arg1[5], const uint6
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664], [0x0 ~> 0x1a666666666664]]
+ * The function fiat_25519_opp negates a field element.
+ *
+ * Postconditions:
+ *   eval out1 mod m = -eval arg1 mod m
+ *
  */
-static void fiat_25519_opp(uint64_t out1[5], const uint64_t arg1[5]) {
-  uint64_t x1 = (UINT64_C(0xfffffffffffda) - (arg1[0]));
-  uint64_t x2 = (UINT64_C(0xffffffffffffe) - (arg1[1]));
-  uint64_t x3 = (UINT64_C(0xffffffffffffe) - (arg1[2]));
-  uint64_t x4 = (UINT64_C(0xffffffffffffe) - (arg1[3]));
-  uint64_t x5 = (UINT64_C(0xffffffffffffe) - (arg1[4]));
+static FIAT_25519_FIAT_INLINE void fiat_25519_opp(fiat_25519_loose_field_element out1, const fiat_25519_tight_field_element arg1) {
+  uint64_t x1;
+  uint64_t x2;
+  uint64_t x3;
+  uint64_t x4;
+  uint64_t x5;
+  x1 = (UINT64_C(0xfffffffffffda) - (arg1[0]));
+  x2 = (UINT64_C(0xffffffffffffe) - (arg1[1]));
+  x3 = (UINT64_C(0xffffffffffffe) - (arg1[2]));
+  x4 = (UINT64_C(0xffffffffffffe) - (arg1[3]));
+  x5 = (UINT64_C(0xffffffffffffe) - (arg1[4]));
   out1[0] = x1;
   out1[1] = x2;
   out1[2] = x3;
@@ -281,227 +474,434 @@ static void fiat_25519_opp(uint64_t out1[5], const uint64_t arg1[5]) {
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc]]
+ * The function fiat_25519_to_bytes serializes a field element to bytes in little-endian order.
+ *
+ * Postconditions:
+ *   out1 = map (λ x, ⌊((eval arg1 mod m) mod 2^(8 * (x + 1))) / 2^(8 * x)⌋) [0..31]
+ *
  * Output Bounds:
  *   out1: [[0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0x7f]]
  */
-static void fiat_25519_to_bytes(uint8_t out1[32], const uint64_t arg1[5]) {
+static FIAT_25519_FIAT_INLINE void fiat_25519_to_bytes(uint8_t out1[32], const fiat_25519_tight_field_element arg1) {
   uint64_t x1;
   fiat_25519_uint1 x2;
-  fiat_25519_subborrowx_u51(&x1, &x2, 0x0, (arg1[0]), UINT64_C(0x7ffffffffffed));
   uint64_t x3;
   fiat_25519_uint1 x4;
-  fiat_25519_subborrowx_u51(&x3, &x4, x2, (arg1[1]), UINT64_C(0x7ffffffffffff));
   uint64_t x5;
   fiat_25519_uint1 x6;
-  fiat_25519_subborrowx_u51(&x5, &x6, x4, (arg1[2]), UINT64_C(0x7ffffffffffff));
   uint64_t x7;
   fiat_25519_uint1 x8;
-  fiat_25519_subborrowx_u51(&x7, &x8, x6, (arg1[3]), UINT64_C(0x7ffffffffffff));
   uint64_t x9;
   fiat_25519_uint1 x10;
-  fiat_25519_subborrowx_u51(&x9, &x10, x8, (arg1[4]), UINT64_C(0x7ffffffffffff));
   uint64_t x11;
-  fiat_25519_cmovznz_u64(&x11, x10, 0x0, UINT64_C(0xffffffffffffffff));
   uint64_t x12;
   fiat_25519_uint1 x13;
-  fiat_25519_addcarryx_u51(&x12, &x13, 0x0, (x11 & UINT64_C(0x7ffffffffffed)), x1);
   uint64_t x14;
   fiat_25519_uint1 x15;
-  fiat_25519_addcarryx_u51(&x14, &x15, x13, (x11 & UINT64_C(0x7ffffffffffff)), x3);
   uint64_t x16;
   fiat_25519_uint1 x17;
-  fiat_25519_addcarryx_u51(&x16, &x17, x15, (x11 & UINT64_C(0x7ffffffffffff)), x5);
   uint64_t x18;
   fiat_25519_uint1 x19;
-  fiat_25519_addcarryx_u51(&x18, &x19, x17, (x11 & UINT64_C(0x7ffffffffffff)), x7);
   uint64_t x20;
   fiat_25519_uint1 x21;
-  fiat_25519_addcarryx_u51(&x20, &x21, x19, (x11 & UINT64_C(0x7ffffffffffff)), x9);
-  uint64_t x22 = (x20 << 4);
-  uint64_t x23 = (x18 * (uint64_t)0x2);
-  uint64_t x24 = (x16 << 6);
-  uint64_t x25 = (x14 << 3);
-  uint64_t x26 = (x12 >> 8);
-  uint8_t x27 = (uint8_t)(x12 & UINT8_C(0xff));
-  uint64_t x28 = (x26 >> 8);
-  uint8_t x29 = (uint8_t)(x26 & UINT8_C(0xff));
-  uint64_t x30 = (x28 >> 8);
-  uint8_t x31 = (uint8_t)(x28 & UINT8_C(0xff));
-  uint64_t x32 = (x30 >> 8);
-  uint8_t x33 = (uint8_t)(x30 & UINT8_C(0xff));
-  uint64_t x34 = (x32 >> 8);
-  uint8_t x35 = (uint8_t)(x32 & UINT8_C(0xff));
-  uint8_t x36 = (uint8_t)(x34 >> 8);
-  uint8_t x37 = (uint8_t)(x34 & UINT8_C(0xff));
-  uint64_t x38 = (x36 + x25);
-  uint64_t x39 = (x38 >> 8);
-  uint8_t x40 = (uint8_t)(x38 & UINT8_C(0xff));
-  uint64_t x41 = (x39 >> 8);
-  uint8_t x42 = (uint8_t)(x39 & UINT8_C(0xff));
-  uint64_t x43 = (x41 >> 8);
-  uint8_t x44 = (uint8_t)(x41 & UINT8_C(0xff));
-  uint64_t x45 = (x43 >> 8);
-  uint8_t x46 = (uint8_t)(x43 & UINT8_C(0xff));
-  uint64_t x47 = (x45 >> 8);
-  uint8_t x48 = (uint8_t)(x45 & UINT8_C(0xff));
-  uint8_t x49 = (uint8_t)(x47 >> 8);
-  uint8_t x50 = (uint8_t)(x47 & UINT8_C(0xff));
-  uint64_t x51 = (x49 + x24);
-  uint64_t x52 = (x51 >> 8);
-  uint8_t x53 = (uint8_t)(x51 & UINT8_C(0xff));
-  uint64_t x54 = (x52 >> 8);
-  uint8_t x55 = (uint8_t)(x52 & UINT8_C(0xff));
-  uint64_t x56 = (x54 >> 8);
-  uint8_t x57 = (uint8_t)(x54 & UINT8_C(0xff));
-  uint64_t x58 = (x56 >> 8);
-  uint8_t x59 = (uint8_t)(x56 & UINT8_C(0xff));
-  uint64_t x60 = (x58 >> 8);
-  uint8_t x61 = (uint8_t)(x58 & UINT8_C(0xff));
-  uint64_t x62 = (x60 >> 8);
-  uint8_t x63 = (uint8_t)(x60 & UINT8_C(0xff));
-  fiat_25519_uint1 x64 = (fiat_25519_uint1)(x62 >> 8);
-  uint8_t x65 = (uint8_t)(x62 & UINT8_C(0xff));
-  uint64_t x66 = (x64 + x23);
-  uint64_t x67 = (x66 >> 8);
-  uint8_t x68 = (uint8_t)(x66 & UINT8_C(0xff));
-  uint64_t x69 = (x67 >> 8);
-  uint8_t x70 = (uint8_t)(x67 & UINT8_C(0xff));
-  uint64_t x71 = (x69 >> 8);
-  uint8_t x72 = (uint8_t)(x69 & UINT8_C(0xff));
-  uint64_t x73 = (x71 >> 8);
-  uint8_t x74 = (uint8_t)(x71 & UINT8_C(0xff));
-  uint64_t x75 = (x73 >> 8);
-  uint8_t x76 = (uint8_t)(x73 & UINT8_C(0xff));
-  uint8_t x77 = (uint8_t)(x75 >> 8);
-  uint8_t x78 = (uint8_t)(x75 & UINT8_C(0xff));
-  uint64_t x79 = (x77 + x22);
-  uint64_t x80 = (x79 >> 8);
-  uint8_t x81 = (uint8_t)(x79 & UINT8_C(0xff));
-  uint64_t x82 = (x80 >> 8);
-  uint8_t x83 = (uint8_t)(x80 & UINT8_C(0xff));
-  uint64_t x84 = (x82 >> 8);
-  uint8_t x85 = (uint8_t)(x82 & UINT8_C(0xff));
-  uint64_t x86 = (x84 >> 8);
-  uint8_t x87 = (uint8_t)(x84 & UINT8_C(0xff));
-  uint64_t x88 = (x86 >> 8);
-  uint8_t x89 = (uint8_t)(x86 & UINT8_C(0xff));
-  uint8_t x90 = (uint8_t)(x88 >> 8);
-  uint8_t x91 = (uint8_t)(x88 & UINT8_C(0xff));
-  out1[0] = x27;
-  out1[1] = x29;
-  out1[2] = x31;
-  out1[3] = x33;
-  out1[4] = x35;
-  out1[5] = x37;
-  out1[6] = x40;
-  out1[7] = x42;
-  out1[8] = x44;
-  out1[9] = x46;
-  out1[10] = x48;
-  out1[11] = x50;
-  out1[12] = x53;
-  out1[13] = x55;
-  out1[14] = x57;
-  out1[15] = x59;
-  out1[16] = x61;
-  out1[17] = x63;
-  out1[18] = x65;
-  out1[19] = x68;
-  out1[20] = x70;
-  out1[21] = x72;
-  out1[22] = x74;
-  out1[23] = x76;
-  out1[24] = x78;
-  out1[25] = x81;
-  out1[26] = x83;
-  out1[27] = x85;
-  out1[28] = x87;
-  out1[29] = x89;
-  out1[30] = x91;
-  out1[31] = x90;
+  uint64_t x22;
+  uint64_t x23;
+  uint64_t x24;
+  uint64_t x25;
+  uint8_t x26;
+  uint64_t x27;
+  uint8_t x28;
+  uint64_t x29;
+  uint8_t x30;
+  uint64_t x31;
+  uint8_t x32;
+  uint64_t x33;
+  uint8_t x34;
+  uint64_t x35;
+  uint8_t x36;
+  uint8_t x37;
+  uint64_t x38;
+  uint8_t x39;
+  uint64_t x40;
+  uint8_t x41;
+  uint64_t x42;
+  uint8_t x43;
+  uint64_t x44;
+  uint8_t x45;
+  uint64_t x46;
+  uint8_t x47;
+  uint64_t x48;
+  uint8_t x49;
+  uint8_t x50;
+  uint64_t x51;
+  uint8_t x52;
+  uint64_t x53;
+  uint8_t x54;
+  uint64_t x55;
+  uint8_t x56;
+  uint64_t x57;
+  uint8_t x58;
+  uint64_t x59;
+  uint8_t x60;
+  uint64_t x61;
+  uint8_t x62;
+  uint64_t x63;
+  uint8_t x64;
+  fiat_25519_uint1 x65;
+  uint64_t x66;
+  uint8_t x67;
+  uint64_t x68;
+  uint8_t x69;
+  uint64_t x70;
+  uint8_t x71;
+  uint64_t x72;
+  uint8_t x73;
+  uint64_t x74;
+  uint8_t x75;
+  uint64_t x76;
+  uint8_t x77;
+  uint8_t x78;
+  uint64_t x79;
+  uint8_t x80;
+  uint64_t x81;
+  uint8_t x82;
+  uint64_t x83;
+  uint8_t x84;
+  uint64_t x85;
+  uint8_t x86;
+  uint64_t x87;
+  uint8_t x88;
+  uint64_t x89;
+  uint8_t x90;
+  uint8_t x91;
+  fiat_25519_subborrowx_u51(&x1, &x2, 0x0, (arg1[0]), UINT64_C(0x7ffffffffffed));
+  fiat_25519_subborrowx_u51(&x3, &x4, x2, (arg1[1]), UINT64_C(0x7ffffffffffff));
+  fiat_25519_subborrowx_u51(&x5, &x6, x4, (arg1[2]), UINT64_C(0x7ffffffffffff));
+  fiat_25519_subborrowx_u51(&x7, &x8, x6, (arg1[3]), UINT64_C(0x7ffffffffffff));
+  fiat_25519_subborrowx_u51(&x9, &x10, x8, (arg1[4]), UINT64_C(0x7ffffffffffff));
+  fiat_25519_cmovznz_u64(&x11, x10, 0x0, UINT64_C(0xffffffffffffffff));
+  fiat_25519_addcarryx_u51(&x12, &x13, 0x0, x1, (x11 & UINT64_C(0x7ffffffffffed)));
+  fiat_25519_addcarryx_u51(&x14, &x15, x13, x3, (x11 & UINT64_C(0x7ffffffffffff)));
+  fiat_25519_addcarryx_u51(&x16, &x17, x15, x5, (x11 & UINT64_C(0x7ffffffffffff)));
+  fiat_25519_addcarryx_u51(&x18, &x19, x17, x7, (x11 & UINT64_C(0x7ffffffffffff)));
+  fiat_25519_addcarryx_u51(&x20, &x21, x19, x9, (x11 & UINT64_C(0x7ffffffffffff)));
+  x22 = (x20 << 4);
+  x23 = (x18 * (uint64_t)0x2);
+  x24 = (x16 << 6);
+  x25 = (x14 << 3);
+  x26 = (uint8_t)(x12 & UINT8_C(0xff));
+  x27 = (x12 >> 8);
+  x28 = (uint8_t)(x27 & UINT8_C(0xff));
+  x29 = (x27 >> 8);
+  x30 = (uint8_t)(x29 & UINT8_C(0xff));
+  x31 = (x29 >> 8);
+  x32 = (uint8_t)(x31 & UINT8_C(0xff));
+  x33 = (x31 >> 8);
+  x34 = (uint8_t)(x33 & UINT8_C(0xff));
+  x35 = (x33 >> 8);
+  x36 = (uint8_t)(x35 & UINT8_C(0xff));
+  x37 = (uint8_t)(x35 >> 8);
+  x38 = (x25 + (uint64_t)x37);
+  x39 = (uint8_t)(x38 & UINT8_C(0xff));
+  x40 = (x38 >> 8);
+  x41 = (uint8_t)(x40 & UINT8_C(0xff));
+  x42 = (x40 >> 8);
+  x43 = (uint8_t)(x42 & UINT8_C(0xff));
+  x44 = (x42 >> 8);
+  x45 = (uint8_t)(x44 & UINT8_C(0xff));
+  x46 = (x44 >> 8);
+  x47 = (uint8_t)(x46 & UINT8_C(0xff));
+  x48 = (x46 >> 8);
+  x49 = (uint8_t)(x48 & UINT8_C(0xff));
+  x50 = (uint8_t)(x48 >> 8);
+  x51 = (x24 + (uint64_t)x50);
+  x52 = (uint8_t)(x51 & UINT8_C(0xff));
+  x53 = (x51 >> 8);
+  x54 = (uint8_t)(x53 & UINT8_C(0xff));
+  x55 = (x53 >> 8);
+  x56 = (uint8_t)(x55 & UINT8_C(0xff));
+  x57 = (x55 >> 8);
+  x58 = (uint8_t)(x57 & UINT8_C(0xff));
+  x59 = (x57 >> 8);
+  x60 = (uint8_t)(x59 & UINT8_C(0xff));
+  x61 = (x59 >> 8);
+  x62 = (uint8_t)(x61 & UINT8_C(0xff));
+  x63 = (x61 >> 8);
+  x64 = (uint8_t)(x63 & UINT8_C(0xff));
+  x65 = (fiat_25519_uint1)(x63 >> 8);
+  x66 = (x23 + (uint64_t)x65);
+  x67 = (uint8_t)(x66 & UINT8_C(0xff));
+  x68 = (x66 >> 8);
+  x69 = (uint8_t)(x68 & UINT8_C(0xff));
+  x70 = (x68 >> 8);
+  x71 = (uint8_t)(x70 & UINT8_C(0xff));
+  x72 = (x70 >> 8);
+  x73 = (uint8_t)(x72 & UINT8_C(0xff));
+  x74 = (x72 >> 8);
+  x75 = (uint8_t)(x74 & UINT8_C(0xff));
+  x76 = (x74 >> 8);
+  x77 = (uint8_t)(x76 & UINT8_C(0xff));
+  x78 = (uint8_t)(x76 >> 8);
+  x79 = (x22 + (uint64_t)x78);
+  x80 = (uint8_t)(x79 & UINT8_C(0xff));
+  x81 = (x79 >> 8);
+  x82 = (uint8_t)(x81 & UINT8_C(0xff));
+  x83 = (x81 >> 8);
+  x84 = (uint8_t)(x83 & UINT8_C(0xff));
+  x85 = (x83 >> 8);
+  x86 = (uint8_t)(x85 & UINT8_C(0xff));
+  x87 = (x85 >> 8);
+  x88 = (uint8_t)(x87 & UINT8_C(0xff));
+  x89 = (x87 >> 8);
+  x90 = (uint8_t)(x89 & UINT8_C(0xff));
+  x91 = (uint8_t)(x89 >> 8);
+  out1[0] = x26;
+  out1[1] = x28;
+  out1[2] = x30;
+  out1[3] = x32;
+  out1[4] = x34;
+  out1[5] = x36;
+  out1[6] = x39;
+  out1[7] = x41;
+  out1[8] = x43;
+  out1[9] = x45;
+  out1[10] = x47;
+  out1[11] = x49;
+  out1[12] = x52;
+  out1[13] = x54;
+  out1[14] = x56;
+  out1[15] = x58;
+  out1[16] = x60;
+  out1[17] = x62;
+  out1[18] = x64;
+  out1[19] = x67;
+  out1[20] = x69;
+  out1[21] = x71;
+  out1[22] = x73;
+  out1[23] = x75;
+  out1[24] = x77;
+  out1[25] = x80;
+  out1[26] = x82;
+  out1[27] = x84;
+  out1[28] = x86;
+  out1[29] = x88;
+  out1[30] = x90;
+  out1[31] = x91;
 }
 
 /*
+ * The function fiat_25519_from_bytes deserializes a field element from bytes in little-endian order.
+ *
+ * Postconditions:
+ *   eval out1 mod m = bytes_eval arg1 mod m
+ *
  * Input Bounds:
  *   arg1: [[0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0x7f]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc], [0x0 ~> 0x8cccccccccccc]]
  */
-static void fiat_25519_from_bytes(uint64_t out1[5], const uint8_t arg1[32]) {
-  uint64_t x1 = ((uint64_t)(arg1[31]) << 44);
-  uint64_t x2 = ((uint64_t)(arg1[30]) << 36);
-  uint64_t x3 = ((uint64_t)(arg1[29]) << 28);
-  uint64_t x4 = ((uint64_t)(arg1[28]) << 20);
-  uint64_t x5 = ((uint64_t)(arg1[27]) << 12);
-  uint64_t x6 = ((uint64_t)(arg1[26]) << 4);
-  uint64_t x7 = ((uint64_t)(arg1[25]) << 47);
-  uint64_t x8 = ((uint64_t)(arg1[24]) << 39);
-  uint64_t x9 = ((uint64_t)(arg1[23]) << 31);
-  uint64_t x10 = ((uint64_t)(arg1[22]) << 23);
-  uint64_t x11 = ((uint64_t)(arg1[21]) << 15);
-  uint64_t x12 = ((uint64_t)(arg1[20]) << 7);
-  uint64_t x13 = ((uint64_t)(arg1[19]) << 50);
-  uint64_t x14 = ((uint64_t)(arg1[18]) << 42);
-  uint64_t x15 = ((uint64_t)(arg1[17]) << 34);
-  uint64_t x16 = ((uint64_t)(arg1[16]) << 26);
-  uint64_t x17 = ((uint64_t)(arg1[15]) << 18);
-  uint64_t x18 = ((uint64_t)(arg1[14]) << 10);
-  uint64_t x19 = ((uint64_t)(arg1[13]) << 2);
-  uint64_t x20 = ((uint64_t)(arg1[12]) << 45);
-  uint64_t x21 = ((uint64_t)(arg1[11]) << 37);
-  uint64_t x22 = ((uint64_t)(arg1[10]) << 29);
-  uint64_t x23 = ((uint64_t)(arg1[9]) << 21);
-  uint64_t x24 = ((uint64_t)(arg1[8]) << 13);
-  uint64_t x25 = ((uint64_t)(arg1[7]) << 5);
-  uint64_t x26 = ((uint64_t)(arg1[6]) << 48);
-  uint64_t x27 = ((uint64_t)(arg1[5]) << 40);
-  uint64_t x28 = ((uint64_t)(arg1[4]) << 32);
-  uint64_t x29 = ((uint64_t)(arg1[3]) << 24);
-  uint64_t x30 = ((uint64_t)(arg1[2]) << 16);
-  uint64_t x31 = ((uint64_t)(arg1[1]) << 8);
-  uint8_t x32 = (arg1[0]);
-  uint64_t x33 = (x32 + (x31 + (x30 + (x29 + (x28 + (x27 + x26))))));
-  uint8_t x34 = (uint8_t)(x33 >> 51);
-  uint64_t x35 = (x33 & UINT64_C(0x7ffffffffffff));
-  uint64_t x36 = (x6 + (x5 + (x4 + (x3 + (x2 + x1)))));
-  uint64_t x37 = (x12 + (x11 + (x10 + (x9 + (x8 + x7)))));
-  uint64_t x38 = (x19 + (x18 + (x17 + (x16 + (x15 + (x14 + x13))))));
-  uint64_t x39 = (x25 + (x24 + (x23 + (x22 + (x21 + x20)))));
-  uint64_t x40 = (x34 + x39);
-  uint8_t x41 = (uint8_t)(x40 >> 51);
-  uint64_t x42 = (x40 & UINT64_C(0x7ffffffffffff));
-  uint64_t x43 = (x41 + x38);
-  uint8_t x44 = (uint8_t)(x43 >> 51);
-  uint64_t x45 = (x43 & UINT64_C(0x7ffffffffffff));
-  uint64_t x46 = (x44 + x37);
-  uint8_t x47 = (uint8_t)(x46 >> 51);
-  uint64_t x48 = (x46 & UINT64_C(0x7ffffffffffff));
-  uint64_t x49 = (x47 + x36);
-  out1[0] = x35;
-  out1[1] = x42;
-  out1[2] = x45;
-  out1[3] = x48;
-  out1[4] = x49;
+static FIAT_25519_FIAT_INLINE void fiat_25519_from_bytes(fiat_25519_tight_field_element out1, const uint8_t arg1[32]) {
+  uint64_t x1;
+  uint64_t x2;
+  uint64_t x3;
+  uint64_t x4;
+  uint64_t x5;
+  uint64_t x6;
+  uint64_t x7;
+  uint64_t x8;
+  uint64_t x9;
+  uint64_t x10;
+  uint64_t x11;
+  uint64_t x12;
+  uint64_t x13;
+  uint64_t x14;
+  uint64_t x15;
+  uint64_t x16;
+  uint64_t x17;
+  uint64_t x18;
+  uint64_t x19;
+  uint64_t x20;
+  uint64_t x21;
+  uint64_t x22;
+  uint64_t x23;
+  uint64_t x24;
+  uint64_t x25;
+  uint64_t x26;
+  uint64_t x27;
+  uint64_t x28;
+  uint64_t x29;
+  uint64_t x30;
+  uint64_t x31;
+  uint8_t x32;
+  uint64_t x33;
+  uint64_t x34;
+  uint64_t x35;
+  uint64_t x36;
+  uint64_t x37;
+  uint64_t x38;
+  uint64_t x39;
+  uint8_t x40;
+  uint64_t x41;
+  uint64_t x42;
+  uint64_t x43;
+  uint64_t x44;
+  uint64_t x45;
+  uint64_t x46;
+  uint64_t x47;
+  uint8_t x48;
+  uint64_t x49;
+  uint64_t x50;
+  uint64_t x51;
+  uint64_t x52;
+  uint64_t x53;
+  uint64_t x54;
+  uint64_t x55;
+  uint64_t x56;
+  uint8_t x57;
+  uint64_t x58;
+  uint64_t x59;
+  uint64_t x60;
+  uint64_t x61;
+  uint64_t x62;
+  uint64_t x63;
+  uint64_t x64;
+  uint8_t x65;
+  uint64_t x66;
+  uint64_t x67;
+  uint64_t x68;
+  uint64_t x69;
+  uint64_t x70;
+  uint64_t x71;
+  x1 = ((uint64_t)(arg1[31]) << 44);
+  x2 = ((uint64_t)(arg1[30]) << 36);
+  x3 = ((uint64_t)(arg1[29]) << 28);
+  x4 = ((uint64_t)(arg1[28]) << 20);
+  x5 = ((uint64_t)(arg1[27]) << 12);
+  x6 = ((uint64_t)(arg1[26]) << 4);
+  x7 = ((uint64_t)(arg1[25]) << 47);
+  x8 = ((uint64_t)(arg1[24]) << 39);
+  x9 = ((uint64_t)(arg1[23]) << 31);
+  x10 = ((uint64_t)(arg1[22]) << 23);
+  x11 = ((uint64_t)(arg1[21]) << 15);
+  x12 = ((uint64_t)(arg1[20]) << 7);
+  x13 = ((uint64_t)(arg1[19]) << 50);
+  x14 = ((uint64_t)(arg1[18]) << 42);
+  x15 = ((uint64_t)(arg1[17]) << 34);
+  x16 = ((uint64_t)(arg1[16]) << 26);
+  x17 = ((uint64_t)(arg1[15]) << 18);
+  x18 = ((uint64_t)(arg1[14]) << 10);
+  x19 = ((uint64_t)(arg1[13]) << 2);
+  x20 = ((uint64_t)(arg1[12]) << 45);
+  x21 = ((uint64_t)(arg1[11]) << 37);
+  x22 = ((uint64_t)(arg1[10]) << 29);
+  x23 = ((uint64_t)(arg1[9]) << 21);
+  x24 = ((uint64_t)(arg1[8]) << 13);
+  x25 = ((uint64_t)(arg1[7]) << 5);
+  x26 = ((uint64_t)(arg1[6]) << 48);
+  x27 = ((uint64_t)(arg1[5]) << 40);
+  x28 = ((uint64_t)(arg1[4]) << 32);
+  x29 = ((uint64_t)(arg1[3]) << 24);
+  x30 = ((uint64_t)(arg1[2]) << 16);
+  x31 = ((uint64_t)(arg1[1]) << 8);
+  x32 = (arg1[0]);
+  x33 = (x31 + (uint64_t)x32);
+  x34 = (x30 + x33);
+  x35 = (x29 + x34);
+  x36 = (x28 + x35);
+  x37 = (x27 + x36);
+  x38 = (x26 + x37);
+  x39 = (x38 & UINT64_C(0x7ffffffffffff));
+  x40 = (uint8_t)(x38 >> 51);
+  x41 = (x25 + (uint64_t)x40);
+  x42 = (x24 + x41);
+  x43 = (x23 + x42);
+  x44 = (x22 + x43);
+  x45 = (x21 + x44);
+  x46 = (x20 + x45);
+  x47 = (x46 & UINT64_C(0x7ffffffffffff));
+  x48 = (uint8_t)(x46 >> 51);
+  x49 = (x19 + (uint64_t)x48);
+  x50 = (x18 + x49);
+  x51 = (x17 + x50);
+  x52 = (x16 + x51);
+  x53 = (x15 + x52);
+  x54 = (x14 + x53);
+  x55 = (x13 + x54);
+  x56 = (x55 & UINT64_C(0x7ffffffffffff));
+  x57 = (uint8_t)(x55 >> 51);
+  x58 = (x12 + (uint64_t)x57);
+  x59 = (x11 + x58);
+  x60 = (x10 + x59);
+  x61 = (x9 + x60);
+  x62 = (x8 + x61);
+  x63 = (x7 + x62);
+  x64 = (x63 & UINT64_C(0x7ffffffffffff));
+  x65 = (uint8_t)(x63 >> 51);
+  x66 = (x6 + (uint64_t)x65);
+  x67 = (x5 + x66);
+  x68 = (x4 + x67);
+  x69 = (x3 + x68);
+  x70 = (x2 + x69);
+  x71 = (x1 + x70);
+  out1[0] = x39;
+  out1[1] = x47;
+  out1[2] = x56;
+  out1[3] = x64;
+  out1[4] = x71;
 }
 
 #else /* defined(BORINGSSL_CURVE25519_64BIT) */
 
-/* Autogenerated */
+/* Autogenerated: 'src/ExtractionOCaml/unsaturated_solinas' --inline --static --use-value-barrier 25519 32 '(auto)' '2^255 - 19' carry_mul carry_square carry add sub opp selectznz to_bytes from_bytes relax carry_scmul121666 */
 /* curve description: 25519 */
-/* requested operations: carry_mul, carry_square, carry_scmul121666, carry, add, sub, opp, selectznz, to_bytes, from_bytes */
-/* n = 10 (from "10") */
-/* s = 0x8000000000000000000000000000000000000000000000000000000000000000 (from "2^255") */
-/* c = [(1, 19)] (from "1,19") */
 /* machine_wordsize = 32 (from "32") */
+/* requested operations: carry_mul, carry_square, carry, add, sub, opp, selectznz, to_bytes, from_bytes, relax, carry_scmul121666 */
+/* n = 10 (from "(auto)") */
+/* s-c = 2^255 - [(1, 19)] (from "2^255 - 19") */
+/* tight_bounds_multiplier = 1 (from "") */
+/*  */
+/* Computed values: */
+/*   carry_chain = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1] */
+/*   eval z = z[0] + (z[1] << 26) + (z[2] << 51) + (z[3] << 77) + (z[4] << 102) + (z[5] << 128) + (z[6] << 153) + (z[7] << 179) + (z[8] << 204) + (z[9] << 230) */
+/*   bytes_eval z = z[0] + (z[1] << 8) + (z[2] << 16) + (z[3] << 24) + (z[4] << 32) + (z[5] << 40) + (z[6] << 48) + (z[7] << 56) + (z[8] << 64) + (z[9] << 72) + (z[10] << 80) + (z[11] << 88) + (z[12] << 96) + (z[13] << 104) + (z[14] << 112) + (z[15] << 120) + (z[16] << 128) + (z[17] << 136) + (z[18] << 144) + (z[19] << 152) + (z[20] << 160) + (z[21] << 168) + (z[22] << 176) + (z[23] << 184) + (z[24] << 192) + (z[25] << 200) + (z[26] << 208) + (z[27] << 216) + (z[28] << 224) + (z[29] << 232) + (z[30] << 240) + (z[31] << 248) */
+/*   balance = [0x7ffffda, 0x3fffffe, 0x7fffffe, 0x3fffffe, 0x7fffffe, 0x3fffffe, 0x7fffffe, 0x3fffffe, 0x7fffffe, 0x3fffffe] */
 
 #include <stdint.h>
 typedef unsigned char fiat_25519_uint1;
 typedef signed char fiat_25519_int1;
+#if defined(__GNUC__) || defined(__clang__)
+#  define FIAT_25519_FIAT_INLINE __inline__
+#else
+#  define FIAT_25519_FIAT_INLINE
+#endif
+
+/* The type fiat_25519_loose_field_element is a field element with loose bounds. */
+/* Bounds: [[0x0 ~> 0xc000000], [0x0 ~> 0x6000000], [0x0 ~> 0xc000000], [0x0 ~> 0x6000000], [0x0 ~> 0xc000000], [0x0 ~> 0x6000000], [0x0 ~> 0xc000000], [0x0 ~> 0x6000000], [0x0 ~> 0xc000000], [0x0 ~> 0x6000000]] */
+typedef uint32_t fiat_25519_loose_field_element[10];
+
+/* The type fiat_25519_tight_field_element is a field element with tight bounds. */
+/* Bounds: [[0x0 ~> 0x4000000], [0x0 ~> 0x2000000], [0x0 ~> 0x4000000], [0x0 ~> 0x2000000], [0x0 ~> 0x4000000], [0x0 ~> 0x2000000], [0x0 ~> 0x4000000], [0x0 ~> 0x2000000], [0x0 ~> 0x4000000], [0x0 ~> 0x2000000]] */
+typedef uint32_t fiat_25519_tight_field_element[10];
+
+#if (-1 & 3) != 3
+#error "This code only works on a two's complement system"
+#endif
+
+#if !defined(FIAT_25519_NO_ASM) && (defined(__GNUC__) || defined(__clang__))
+static __inline__ uint32_t fiat_25519_value_barrier_u32(uint32_t a) {
+  __asm__("" : "+r"(a) : /* no inputs */);
+  return a;
+}
+#else
+#  define fiat_25519_value_barrier_u32(x) (x)
+#endif
 
 
 /*
+ * The function fiat_25519_addcarryx_u26 is an addition with carry.
+ *
+ * Postconditions:
+ *   out1 = (arg1 + arg2 + arg3) mod 2^26
+ *   out2 = ⌊(arg1 + arg2 + arg3) / 2^26⌋
+ *
  * Input Bounds:
  *   arg1: [0x0 ~> 0x1]
  *   arg2: [0x0 ~> 0x3ffffff]
@@ -510,15 +910,24 @@ typedef signed char fiat_25519_int1;
  *   out1: [0x0 ~> 0x3ffffff]
  *   out2: [0x0 ~> 0x1]
  */
-static void fiat_25519_addcarryx_u26(uint32_t* out1, fiat_25519_uint1* out2, fiat_25519_uint1 arg1, uint32_t arg2, uint32_t arg3) {
-  uint32_t x1 = ((arg1 + arg2) + arg3);
-  uint32_t x2 = (x1 & UINT32_C(0x3ffffff));
-  fiat_25519_uint1 x3 = (fiat_25519_uint1)(x1 >> 26);
+static FIAT_25519_FIAT_INLINE void fiat_25519_addcarryx_u26(uint32_t* out1, fiat_25519_uint1* out2, fiat_25519_uint1 arg1, uint32_t arg2, uint32_t arg3) {
+  uint32_t x1;
+  uint32_t x2;
+  fiat_25519_uint1 x3;
+  x1 = ((arg1 + arg2) + arg3);
+  x2 = (x1 & UINT32_C(0x3ffffff));
+  x3 = (fiat_25519_uint1)(x1 >> 26);
   *out1 = x2;
   *out2 = x3;
 }
 
 /*
+ * The function fiat_25519_subborrowx_u26 is a subtraction with borrow.
+ *
+ * Postconditions:
+ *   out1 = (-arg1 + arg2 + -arg3) mod 2^26
+ *   out2 = -⌊(-arg1 + arg2 + -arg3) / 2^26⌋
+ *
  * Input Bounds:
  *   arg1: [0x0 ~> 0x1]
  *   arg2: [0x0 ~> 0x3ffffff]
@@ -527,15 +936,24 @@ static void fiat_25519_addcarryx_u26(uint32_t* out1, fiat_25519_uint1* out2, fia
  *   out1: [0x0 ~> 0x3ffffff]
  *   out2: [0x0 ~> 0x1]
  */
-static void fiat_25519_subborrowx_u26(uint32_t* out1, fiat_25519_uint1* out2, fiat_25519_uint1 arg1, uint32_t arg2, uint32_t arg3) {
-  int32_t x1 = ((int32_t)(arg2 - arg1) - (int32_t)arg3);
-  fiat_25519_int1 x2 = (fiat_25519_int1)(x1 >> 26);
-  uint32_t x3 = (x1 & UINT32_C(0x3ffffff));
+static FIAT_25519_FIAT_INLINE void fiat_25519_subborrowx_u26(uint32_t* out1, fiat_25519_uint1* out2, fiat_25519_uint1 arg1, uint32_t arg2, uint32_t arg3) {
+  int32_t x1;
+  fiat_25519_int1 x2;
+  uint32_t x3;
+  x1 = ((int32_t)(arg2 - arg1) - (int32_t)arg3);
+  x2 = (fiat_25519_int1)(x1 >> 26);
+  x3 = (x1 & UINT32_C(0x3ffffff));
   *out1 = x3;
   *out2 = (fiat_25519_uint1)(0x0 - x2);
 }
 
 /*
+ * The function fiat_25519_addcarryx_u25 is an addition with carry.
+ *
+ * Postconditions:
+ *   out1 = (arg1 + arg2 + arg3) mod 2^25
+ *   out2 = ⌊(arg1 + arg2 + arg3) / 2^25⌋
+ *
  * Input Bounds:
  *   arg1: [0x0 ~> 0x1]
  *   arg2: [0x0 ~> 0x1ffffff]
@@ -544,15 +962,24 @@ static void fiat_25519_subborrowx_u26(uint32_t* out1, fiat_25519_uint1* out2, fi
  *   out1: [0x0 ~> 0x1ffffff]
  *   out2: [0x0 ~> 0x1]
  */
-static void fiat_25519_addcarryx_u25(uint32_t* out1, fiat_25519_uint1* out2, fiat_25519_uint1 arg1, uint32_t arg2, uint32_t arg3) {
-  uint32_t x1 = ((arg1 + arg2) + arg3);
-  uint32_t x2 = (x1 & UINT32_C(0x1ffffff));
-  fiat_25519_uint1 x3 = (fiat_25519_uint1)(x1 >> 25);
+static FIAT_25519_FIAT_INLINE void fiat_25519_addcarryx_u25(uint32_t* out1, fiat_25519_uint1* out2, fiat_25519_uint1 arg1, uint32_t arg2, uint32_t arg3) {
+  uint32_t x1;
+  uint32_t x2;
+  fiat_25519_uint1 x3;
+  x1 = ((arg1 + arg2) + arg3);
+  x2 = (x1 & UINT32_C(0x1ffffff));
+  x3 = (fiat_25519_uint1)(x1 >> 25);
   *out1 = x2;
   *out2 = x3;
 }
 
 /*
+ * The function fiat_25519_subborrowx_u25 is a subtraction with borrow.
+ *
+ * Postconditions:
+ *   out1 = (-arg1 + arg2 + -arg3) mod 2^25
+ *   out2 = -⌊(-arg1 + arg2 + -arg3) / 2^25⌋
+ *
  * Input Bounds:
  *   arg1: [0x0 ~> 0x1]
  *   arg2: [0x0 ~> 0x1ffffff]
@@ -561,15 +988,23 @@ static void fiat_25519_addcarryx_u25(uint32_t* out1, fiat_25519_uint1* out2, fia
  *   out1: [0x0 ~> 0x1ffffff]
  *   out2: [0x0 ~> 0x1]
  */
-static void fiat_25519_subborrowx_u25(uint32_t* out1, fiat_25519_uint1* out2, fiat_25519_uint1 arg1, uint32_t arg2, uint32_t arg3) {
-  int32_t x1 = ((int32_t)(arg2 - arg1) - (int32_t)arg3);
-  fiat_25519_int1 x2 = (fiat_25519_int1)(x1 >> 25);
-  uint32_t x3 = (x1 & UINT32_C(0x1ffffff));
+static FIAT_25519_FIAT_INLINE void fiat_25519_subborrowx_u25(uint32_t* out1, fiat_25519_uint1* out2, fiat_25519_uint1 arg1, uint32_t arg2, uint32_t arg3) {
+  int32_t x1;
+  fiat_25519_int1 x2;
+  uint32_t x3;
+  x1 = ((int32_t)(arg2 - arg1) - (int32_t)arg3);
+  x2 = (fiat_25519_int1)(x1 >> 25);
+  x3 = (x1 & UINT32_C(0x1ffffff));
   *out1 = x3;
   *out2 = (fiat_25519_uint1)(0x0 - x2);
 }
 
 /*
+ * The function fiat_25519_cmovznz_u32 is a single-word conditional move.
+ *
+ * Postconditions:
+ *   out1 = (if arg1 = 0 then arg2 else arg3)
+ *
  * Input Bounds:
  *   arg1: [0x0 ~> 0x1]
  *   arg2: [0x0 ~> 0xffffffff]
@@ -577,168 +1012,318 @@ static void fiat_25519_subborrowx_u25(uint32_t* out1, fiat_25519_uint1* out2, fi
  * Output Bounds:
  *   out1: [0x0 ~> 0xffffffff]
  */
-static void fiat_25519_cmovznz_u32(uint32_t* out1, fiat_25519_uint1 arg1, uint32_t arg2, uint32_t arg3) {
-  fiat_25519_uint1 x1 = (!(!arg1));
-  uint32_t x2 = ((fiat_25519_int1)(0x0 - x1) & UINT32_C(0xffffffff));
-  uint32_t x3 = ((x2 & arg3) | ((~x2) & arg2));
+static FIAT_25519_FIAT_INLINE void fiat_25519_cmovznz_u32(uint32_t* out1, fiat_25519_uint1 arg1, uint32_t arg2, uint32_t arg3) {
+  fiat_25519_uint1 x1;
+  uint32_t x2;
+  uint32_t x3;
+  x1 = (!(!arg1));
+  x2 = ((fiat_25519_int1)(0x0 - x1) & UINT32_C(0xffffffff));
+  x3 = ((fiat_25519_value_barrier_u32(x2) & arg3) | (fiat_25519_value_barrier_u32((~x2)) & arg2));
   *out1 = x3;
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999]]
- *   arg2: [[0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333]]
+ * The function fiat_25519_carry_mul multiplies two field elements and reduces the result.
+ *
+ * Postconditions:
+ *   eval out1 mod m = (eval arg1 * eval arg2) mod m
+ *
  */
-static void fiat_25519_carry_mul(uint32_t out1[10], const uint32_t arg1[10], const uint32_t arg2[10]) {
-  uint64_t x1 = ((uint64_t)(arg1[9]) * ((arg2[9]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x2 = ((uint64_t)(arg1[9]) * ((arg2[8]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x3 = ((uint64_t)(arg1[9]) * ((arg2[7]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x4 = ((uint64_t)(arg1[9]) * ((arg2[6]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x5 = ((uint64_t)(arg1[9]) * ((arg2[5]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x6 = ((uint64_t)(arg1[9]) * ((arg2[4]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x7 = ((uint64_t)(arg1[9]) * ((arg2[3]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x8 = ((uint64_t)(arg1[9]) * ((arg2[2]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x9 = ((uint64_t)(arg1[9]) * ((arg2[1]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x10 = ((uint64_t)(arg1[8]) * ((arg2[9]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x11 = ((uint64_t)(arg1[8]) * ((arg2[8]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x12 = ((uint64_t)(arg1[8]) * ((arg2[7]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x13 = ((uint64_t)(arg1[8]) * ((arg2[6]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x14 = ((uint64_t)(arg1[8]) * ((arg2[5]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x15 = ((uint64_t)(arg1[8]) * ((arg2[4]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x16 = ((uint64_t)(arg1[8]) * ((arg2[3]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x17 = ((uint64_t)(arg1[8]) * ((arg2[2]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x18 = ((uint64_t)(arg1[7]) * ((arg2[9]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x19 = ((uint64_t)(arg1[7]) * ((arg2[8]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x20 = ((uint64_t)(arg1[7]) * ((arg2[7]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x21 = ((uint64_t)(arg1[7]) * ((arg2[6]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x22 = ((uint64_t)(arg1[7]) * ((arg2[5]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x23 = ((uint64_t)(arg1[7]) * ((arg2[4]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x24 = ((uint64_t)(arg1[7]) * ((arg2[3]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x25 = ((uint64_t)(arg1[6]) * ((arg2[9]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x26 = ((uint64_t)(arg1[6]) * ((arg2[8]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x27 = ((uint64_t)(arg1[6]) * ((arg2[7]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x28 = ((uint64_t)(arg1[6]) * ((arg2[6]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x29 = ((uint64_t)(arg1[6]) * ((arg2[5]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x30 = ((uint64_t)(arg1[6]) * ((arg2[4]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x31 = ((uint64_t)(arg1[5]) * ((arg2[9]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x32 = ((uint64_t)(arg1[5]) * ((arg2[8]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x33 = ((uint64_t)(arg1[5]) * ((arg2[7]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x34 = ((uint64_t)(arg1[5]) * ((arg2[6]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x35 = ((uint64_t)(arg1[5]) * ((arg2[5]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x36 = ((uint64_t)(arg1[4]) * ((arg2[9]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x37 = ((uint64_t)(arg1[4]) * ((arg2[8]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x38 = ((uint64_t)(arg1[4]) * ((arg2[7]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x39 = ((uint64_t)(arg1[4]) * ((arg2[6]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x40 = ((uint64_t)(arg1[3]) * ((arg2[9]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x41 = ((uint64_t)(arg1[3]) * ((arg2[8]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x42 = ((uint64_t)(arg1[3]) * ((arg2[7]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x43 = ((uint64_t)(arg1[2]) * ((arg2[9]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x44 = ((uint64_t)(arg1[2]) * ((arg2[8]) * (uint32_t)UINT8_C(0x13)));
-  uint64_t x45 = ((uint64_t)(arg1[1]) * ((arg2[9]) * ((uint32_t)0x2 * UINT8_C(0x13))));
-  uint64_t x46 = ((uint64_t)(arg1[9]) * (arg2[0]));
-  uint64_t x47 = ((uint64_t)(arg1[8]) * (arg2[1]));
-  uint64_t x48 = ((uint64_t)(arg1[8]) * (arg2[0]));
-  uint64_t x49 = ((uint64_t)(arg1[7]) * (arg2[2]));
-  uint64_t x50 = ((uint64_t)(arg1[7]) * ((arg2[1]) * (uint32_t)0x2));
-  uint64_t x51 = ((uint64_t)(arg1[7]) * (arg2[0]));
-  uint64_t x52 = ((uint64_t)(arg1[6]) * (arg2[3]));
-  uint64_t x53 = ((uint64_t)(arg1[6]) * (arg2[2]));
-  uint64_t x54 = ((uint64_t)(arg1[6]) * (arg2[1]));
-  uint64_t x55 = ((uint64_t)(arg1[6]) * (arg2[0]));
-  uint64_t x56 = ((uint64_t)(arg1[5]) * (arg2[4]));
-  uint64_t x57 = ((uint64_t)(arg1[5]) * ((arg2[3]) * (uint32_t)0x2));
-  uint64_t x58 = ((uint64_t)(arg1[5]) * (arg2[2]));
-  uint64_t x59 = ((uint64_t)(arg1[5]) * ((arg2[1]) * (uint32_t)0x2));
-  uint64_t x60 = ((uint64_t)(arg1[5]) * (arg2[0]));
-  uint64_t x61 = ((uint64_t)(arg1[4]) * (arg2[5]));
-  uint64_t x62 = ((uint64_t)(arg1[4]) * (arg2[4]));
-  uint64_t x63 = ((uint64_t)(arg1[4]) * (arg2[3]));
-  uint64_t x64 = ((uint64_t)(arg1[4]) * (arg2[2]));
-  uint64_t x65 = ((uint64_t)(arg1[4]) * (arg2[1]));
-  uint64_t x66 = ((uint64_t)(arg1[4]) * (arg2[0]));
-  uint64_t x67 = ((uint64_t)(arg1[3]) * (arg2[6]));
-  uint64_t x68 = ((uint64_t)(arg1[3]) * ((arg2[5]) * (uint32_t)0x2));
-  uint64_t x69 = ((uint64_t)(arg1[3]) * (arg2[4]));
-  uint64_t x70 = ((uint64_t)(arg1[3]) * ((arg2[3]) * (uint32_t)0x2));
-  uint64_t x71 = ((uint64_t)(arg1[3]) * (arg2[2]));
-  uint64_t x72 = ((uint64_t)(arg1[3]) * ((arg2[1]) * (uint32_t)0x2));
-  uint64_t x73 = ((uint64_t)(arg1[3]) * (arg2[0]));
-  uint64_t x74 = ((uint64_t)(arg1[2]) * (arg2[7]));
-  uint64_t x75 = ((uint64_t)(arg1[2]) * (arg2[6]));
-  uint64_t x76 = ((uint64_t)(arg1[2]) * (arg2[5]));
-  uint64_t x77 = ((uint64_t)(arg1[2]) * (arg2[4]));
-  uint64_t x78 = ((uint64_t)(arg1[2]) * (arg2[3]));
-  uint64_t x79 = ((uint64_t)(arg1[2]) * (arg2[2]));
-  uint64_t x80 = ((uint64_t)(arg1[2]) * (arg2[1]));
-  uint64_t x81 = ((uint64_t)(arg1[2]) * (arg2[0]));
-  uint64_t x82 = ((uint64_t)(arg1[1]) * (arg2[8]));
-  uint64_t x83 = ((uint64_t)(arg1[1]) * ((arg2[7]) * (uint32_t)0x2));
-  uint64_t x84 = ((uint64_t)(arg1[1]) * (arg2[6]));
-  uint64_t x85 = ((uint64_t)(arg1[1]) * ((arg2[5]) * (uint32_t)0x2));
-  uint64_t x86 = ((uint64_t)(arg1[1]) * (arg2[4]));
-  uint64_t x87 = ((uint64_t)(arg1[1]) * ((arg2[3]) * (uint32_t)0x2));
-  uint64_t x88 = ((uint64_t)(arg1[1]) * (arg2[2]));
-  uint64_t x89 = ((uint64_t)(arg1[1]) * ((arg2[1]) * (uint32_t)0x2));
-  uint64_t x90 = ((uint64_t)(arg1[1]) * (arg2[0]));
-  uint64_t x91 = ((uint64_t)(arg1[0]) * (arg2[9]));
-  uint64_t x92 = ((uint64_t)(arg1[0]) * (arg2[8]));
-  uint64_t x93 = ((uint64_t)(arg1[0]) * (arg2[7]));
-  uint64_t x94 = ((uint64_t)(arg1[0]) * (arg2[6]));
-  uint64_t x95 = ((uint64_t)(arg1[0]) * (arg2[5]));
-  uint64_t x96 = ((uint64_t)(arg1[0]) * (arg2[4]));
-  uint64_t x97 = ((uint64_t)(arg1[0]) * (arg2[3]));
-  uint64_t x98 = ((uint64_t)(arg1[0]) * (arg2[2]));
-  uint64_t x99 = ((uint64_t)(arg1[0]) * (arg2[1]));
-  uint64_t x100 = ((uint64_t)(arg1[0]) * (arg2[0]));
-  uint64_t x101 = (x100 + (x45 + (x44 + (x42 + (x39 + (x35 + (x30 + (x24 + (x17 + x9)))))))));
-  uint64_t x102 = (x101 >> 26);
-  uint32_t x103 = (uint32_t)(x101 & UINT32_C(0x3ffffff));
-  uint64_t x104 = (x91 + (x82 + (x74 + (x67 + (x61 + (x56 + (x52 + (x49 + (x47 + x46)))))))));
-  uint64_t x105 = (x92 + (x83 + (x75 + (x68 + (x62 + (x57 + (x53 + (x50 + (x48 + x1)))))))));
-  uint64_t x106 = (x93 + (x84 + (x76 + (x69 + (x63 + (x58 + (x54 + (x51 + (x10 + x2)))))))));
-  uint64_t x107 = (x94 + (x85 + (x77 + (x70 + (x64 + (x59 + (x55 + (x18 + (x11 + x3)))))))));
-  uint64_t x108 = (x95 + (x86 + (x78 + (x71 + (x65 + (x60 + (x25 + (x19 + (x12 + x4)))))))));
-  uint64_t x109 = (x96 + (x87 + (x79 + (x72 + (x66 + (x31 + (x26 + (x20 + (x13 + x5)))))))));
-  uint64_t x110 = (x97 + (x88 + (x80 + (x73 + (x36 + (x32 + (x27 + (x21 + (x14 + x6)))))))));
-  uint64_t x111 = (x98 + (x89 + (x81 + (x40 + (x37 + (x33 + (x28 + (x22 + (x15 + x7)))))))));
-  uint64_t x112 = (x99 + (x90 + (x43 + (x41 + (x38 + (x34 + (x29 + (x23 + (x16 + x8)))))))));
-  uint64_t x113 = (x102 + x112);
-  uint64_t x114 = (x113 >> 25);
-  uint32_t x115 = (uint32_t)(x113 & UINT32_C(0x1ffffff));
-  uint64_t x116 = (x114 + x111);
-  uint64_t x117 = (x116 >> 26);
-  uint32_t x118 = (uint32_t)(x116 & UINT32_C(0x3ffffff));
-  uint64_t x119 = (x117 + x110);
-  uint64_t x120 = (x119 >> 25);
-  uint32_t x121 = (uint32_t)(x119 & UINT32_C(0x1ffffff));
-  uint64_t x122 = (x120 + x109);
-  uint64_t x123 = (x122 >> 26);
-  uint32_t x124 = (uint32_t)(x122 & UINT32_C(0x3ffffff));
-  uint64_t x125 = (x123 + x108);
-  uint64_t x126 = (x125 >> 25);
-  uint32_t x127 = (uint32_t)(x125 & UINT32_C(0x1ffffff));
-  uint64_t x128 = (x126 + x107);
-  uint64_t x129 = (x128 >> 26);
-  uint32_t x130 = (uint32_t)(x128 & UINT32_C(0x3ffffff));
-  uint64_t x131 = (x129 + x106);
-  uint64_t x132 = (x131 >> 25);
-  uint32_t x133 = (uint32_t)(x131 & UINT32_C(0x1ffffff));
-  uint64_t x134 = (x132 + x105);
-  uint64_t x135 = (x134 >> 26);
-  uint32_t x136 = (uint32_t)(x134 & UINT32_C(0x3ffffff));
-  uint64_t x137 = (x135 + x104);
-  uint64_t x138 = (x137 >> 25);
-  uint32_t x139 = (uint32_t)(x137 & UINT32_C(0x1ffffff));
-  uint64_t x140 = (x138 * (uint64_t)UINT8_C(0x13));
-  uint64_t x141 = (x103 + x140);
-  uint32_t x142 = (uint32_t)(x141 >> 26);
-  uint32_t x143 = (uint32_t)(x141 & UINT32_C(0x3ffffff));
-  uint32_t x144 = (x142 + x115);
-  uint32_t x145 = (x144 >> 25);
-  uint32_t x146 = (x144 & UINT32_C(0x1ffffff));
-  uint32_t x147 = (x145 + x118);
+static FIAT_25519_FIAT_INLINE void fiat_25519_carry_mul(fiat_25519_tight_field_element out1, const fiat_25519_loose_field_element arg1, const fiat_25519_loose_field_element arg2) {
+  uint64_t x1;
+  uint64_t x2;
+  uint64_t x3;
+  uint64_t x4;
+  uint64_t x5;
+  uint64_t x6;
+  uint64_t x7;
+  uint64_t x8;
+  uint64_t x9;
+  uint64_t x10;
+  uint64_t x11;
+  uint64_t x12;
+  uint64_t x13;
+  uint64_t x14;
+  uint64_t x15;
+  uint64_t x16;
+  uint64_t x17;
+  uint64_t x18;
+  uint64_t x19;
+  uint64_t x20;
+  uint64_t x21;
+  uint64_t x22;
+  uint64_t x23;
+  uint64_t x24;
+  uint64_t x25;
+  uint64_t x26;
+  uint64_t x27;
+  uint64_t x28;
+  uint64_t x29;
+  uint64_t x30;
+  uint64_t x31;
+  uint64_t x32;
+  uint64_t x33;
+  uint64_t x34;
+  uint64_t x35;
+  uint64_t x36;
+  uint64_t x37;
+  uint64_t x38;
+  uint64_t x39;
+  uint64_t x40;
+  uint64_t x41;
+  uint64_t x42;
+  uint64_t x43;
+  uint64_t x44;
+  uint64_t x45;
+  uint64_t x46;
+  uint64_t x47;
+  uint64_t x48;
+  uint64_t x49;
+  uint64_t x50;
+  uint64_t x51;
+  uint64_t x52;
+  uint64_t x53;
+  uint64_t x54;
+  uint64_t x55;
+  uint64_t x56;
+  uint64_t x57;
+  uint64_t x58;
+  uint64_t x59;
+  uint64_t x60;
+  uint64_t x61;
+  uint64_t x62;
+  uint64_t x63;
+  uint64_t x64;
+  uint64_t x65;
+  uint64_t x66;
+  uint64_t x67;
+  uint64_t x68;
+  uint64_t x69;
+  uint64_t x70;
+  uint64_t x71;
+  uint64_t x72;
+  uint64_t x73;
+  uint64_t x74;
+  uint64_t x75;
+  uint64_t x76;
+  uint64_t x77;
+  uint64_t x78;
+  uint64_t x79;
+  uint64_t x80;
+  uint64_t x81;
+  uint64_t x82;
+  uint64_t x83;
+  uint64_t x84;
+  uint64_t x85;
+  uint64_t x86;
+  uint64_t x87;
+  uint64_t x88;
+  uint64_t x89;
+  uint64_t x90;
+  uint64_t x91;
+  uint64_t x92;
+  uint64_t x93;
+  uint64_t x94;
+  uint64_t x95;
+  uint64_t x96;
+  uint64_t x97;
+  uint64_t x98;
+  uint64_t x99;
+  uint64_t x100;
+  uint64_t x101;
+  uint64_t x102;
+  uint32_t x103;
+  uint64_t x104;
+  uint64_t x105;
+  uint64_t x106;
+  uint64_t x107;
+  uint64_t x108;
+  uint64_t x109;
+  uint64_t x110;
+  uint64_t x111;
+  uint64_t x112;
+  uint64_t x113;
+  uint64_t x114;
+  uint32_t x115;
+  uint64_t x116;
+  uint64_t x117;
+  uint32_t x118;
+  uint64_t x119;
+  uint64_t x120;
+  uint32_t x121;
+  uint64_t x122;
+  uint64_t x123;
+  uint32_t x124;
+  uint64_t x125;
+  uint64_t x126;
+  uint32_t x127;
+  uint64_t x128;
+  uint64_t x129;
+  uint32_t x130;
+  uint64_t x131;
+  uint64_t x132;
+  uint32_t x133;
+  uint64_t x134;
+  uint64_t x135;
+  uint32_t x136;
+  uint64_t x137;
+  uint64_t x138;
+  uint32_t x139;
+  uint64_t x140;
+  uint64_t x141;
+  uint32_t x142;
+  uint32_t x143;
+  uint32_t x144;
+  fiat_25519_uint1 x145;
+  uint32_t x146;
+  uint32_t x147;
+  x1 = ((uint64_t)(arg1[9]) * ((arg2[9]) * UINT8_C(0x26)));
+  x2 = ((uint64_t)(arg1[9]) * ((arg2[8]) * UINT8_C(0x13)));
+  x3 = ((uint64_t)(arg1[9]) * ((arg2[7]) * UINT8_C(0x26)));
+  x4 = ((uint64_t)(arg1[9]) * ((arg2[6]) * UINT8_C(0x13)));
+  x5 = ((uint64_t)(arg1[9]) * ((arg2[5]) * UINT8_C(0x26)));
+  x6 = ((uint64_t)(arg1[9]) * ((arg2[4]) * UINT8_C(0x13)));
+  x7 = ((uint64_t)(arg1[9]) * ((arg2[3]) * UINT8_C(0x26)));
+  x8 = ((uint64_t)(arg1[9]) * ((arg2[2]) * UINT8_C(0x13)));
+  x9 = ((uint64_t)(arg1[9]) * ((arg2[1]) * UINT8_C(0x26)));
+  x10 = ((uint64_t)(arg1[8]) * ((arg2[9]) * UINT8_C(0x13)));
+  x11 = ((uint64_t)(arg1[8]) * ((arg2[8]) * UINT8_C(0x13)));
+  x12 = ((uint64_t)(arg1[8]) * ((arg2[7]) * UINT8_C(0x13)));
+  x13 = ((uint64_t)(arg1[8]) * ((arg2[6]) * UINT8_C(0x13)));
+  x14 = ((uint64_t)(arg1[8]) * ((arg2[5]) * UINT8_C(0x13)));
+  x15 = ((uint64_t)(arg1[8]) * ((arg2[4]) * UINT8_C(0x13)));
+  x16 = ((uint64_t)(arg1[8]) * ((arg2[3]) * UINT8_C(0x13)));
+  x17 = ((uint64_t)(arg1[8]) * ((arg2[2]) * UINT8_C(0x13)));
+  x18 = ((uint64_t)(arg1[7]) * ((arg2[9]) * UINT8_C(0x26)));
+  x19 = ((uint64_t)(arg1[7]) * ((arg2[8]) * UINT8_C(0x13)));
+  x20 = ((uint64_t)(arg1[7]) * ((arg2[7]) * UINT8_C(0x26)));
+  x21 = ((uint64_t)(arg1[7]) * ((arg2[6]) * UINT8_C(0x13)));
+  x22 = ((uint64_t)(arg1[7]) * ((arg2[5]) * UINT8_C(0x26)));
+  x23 = ((uint64_t)(arg1[7]) * ((arg2[4]) * UINT8_C(0x13)));
+  x24 = ((uint64_t)(arg1[7]) * ((arg2[3]) * UINT8_C(0x26)));
+  x25 = ((uint64_t)(arg1[6]) * ((arg2[9]) * UINT8_C(0x13)));
+  x26 = ((uint64_t)(arg1[6]) * ((arg2[8]) * UINT8_C(0x13)));
+  x27 = ((uint64_t)(arg1[6]) * ((arg2[7]) * UINT8_C(0x13)));
+  x28 = ((uint64_t)(arg1[6]) * ((arg2[6]) * UINT8_C(0x13)));
+  x29 = ((uint64_t)(arg1[6]) * ((arg2[5]) * UINT8_C(0x13)));
+  x30 = ((uint64_t)(arg1[6]) * ((arg2[4]) * UINT8_C(0x13)));
+  x31 = ((uint64_t)(arg1[5]) * ((arg2[9]) * UINT8_C(0x26)));
+  x32 = ((uint64_t)(arg1[5]) * ((arg2[8]) * UINT8_C(0x13)));
+  x33 = ((uint64_t)(arg1[5]) * ((arg2[7]) * UINT8_C(0x26)));
+  x34 = ((uint64_t)(arg1[5]) * ((arg2[6]) * UINT8_C(0x13)));
+  x35 = ((uint64_t)(arg1[5]) * ((arg2[5]) * UINT8_C(0x26)));
+  x36 = ((uint64_t)(arg1[4]) * ((arg2[9]) * UINT8_C(0x13)));
+  x37 = ((uint64_t)(arg1[4]) * ((arg2[8]) * UINT8_C(0x13)));
+  x38 = ((uint64_t)(arg1[4]) * ((arg2[7]) * UINT8_C(0x13)));
+  x39 = ((uint64_t)(arg1[4]) * ((arg2[6]) * UINT8_C(0x13)));
+  x40 = ((uint64_t)(arg1[3]) * ((arg2[9]) * UINT8_C(0x26)));
+  x41 = ((uint64_t)(arg1[3]) * ((arg2[8]) * UINT8_C(0x13)));
+  x42 = ((uint64_t)(arg1[3]) * ((arg2[7]) * UINT8_C(0x26)));
+  x43 = ((uint64_t)(arg1[2]) * ((arg2[9]) * UINT8_C(0x13)));
+  x44 = ((uint64_t)(arg1[2]) * ((arg2[8]) * UINT8_C(0x13)));
+  x45 = ((uint64_t)(arg1[1]) * ((arg2[9]) * UINT8_C(0x26)));
+  x46 = ((uint64_t)(arg1[9]) * (arg2[0]));
+  x47 = ((uint64_t)(arg1[8]) * (arg2[1]));
+  x48 = ((uint64_t)(arg1[8]) * (arg2[0]));
+  x49 = ((uint64_t)(arg1[7]) * (arg2[2]));
+  x50 = ((uint64_t)(arg1[7]) * ((arg2[1]) * 0x2));
+  x51 = ((uint64_t)(arg1[7]) * (arg2[0]));
+  x52 = ((uint64_t)(arg1[6]) * (arg2[3]));
+  x53 = ((uint64_t)(arg1[6]) * (arg2[2]));
+  x54 = ((uint64_t)(arg1[6]) * (arg2[1]));
+  x55 = ((uint64_t)(arg1[6]) * (arg2[0]));
+  x56 = ((uint64_t)(arg1[5]) * (arg2[4]));
+  x57 = ((uint64_t)(arg1[5]) * ((arg2[3]) * 0x2));
+  x58 = ((uint64_t)(arg1[5]) * (arg2[2]));
+  x59 = ((uint64_t)(arg1[5]) * ((arg2[1]) * 0x2));
+  x60 = ((uint64_t)(arg1[5]) * (arg2[0]));
+  x61 = ((uint64_t)(arg1[4]) * (arg2[5]));
+  x62 = ((uint64_t)(arg1[4]) * (arg2[4]));
+  x63 = ((uint64_t)(arg1[4]) * (arg2[3]));
+  x64 = ((uint64_t)(arg1[4]) * (arg2[2]));
+  x65 = ((uint64_t)(arg1[4]) * (arg2[1]));
+  x66 = ((uint64_t)(arg1[4]) * (arg2[0]));
+  x67 = ((uint64_t)(arg1[3]) * (arg2[6]));
+  x68 = ((uint64_t)(arg1[3]) * ((arg2[5]) * 0x2));
+  x69 = ((uint64_t)(arg1[3]) * (arg2[4]));
+  x70 = ((uint64_t)(arg1[3]) * ((arg2[3]) * 0x2));
+  x71 = ((uint64_t)(arg1[3]) * (arg2[2]));
+  x72 = ((uint64_t)(arg1[3]) * ((arg2[1]) * 0x2));
+  x73 = ((uint64_t)(arg1[3]) * (arg2[0]));
+  x74 = ((uint64_t)(arg1[2]) * (arg2[7]));
+  x75 = ((uint64_t)(arg1[2]) * (arg2[6]));
+  x76 = ((uint64_t)(arg1[2]) * (arg2[5]));
+  x77 = ((uint64_t)(arg1[2]) * (arg2[4]));
+  x78 = ((uint64_t)(arg1[2]) * (arg2[3]));
+  x79 = ((uint64_t)(arg1[2]) * (arg2[2]));
+  x80 = ((uint64_t)(arg1[2]) * (arg2[1]));
+  x81 = ((uint64_t)(arg1[2]) * (arg2[0]));
+  x82 = ((uint64_t)(arg1[1]) * (arg2[8]));
+  x83 = ((uint64_t)(arg1[1]) * ((arg2[7]) * 0x2));
+  x84 = ((uint64_t)(arg1[1]) * (arg2[6]));
+  x85 = ((uint64_t)(arg1[1]) * ((arg2[5]) * 0x2));
+  x86 = ((uint64_t)(arg1[1]) * (arg2[4]));
+  x87 = ((uint64_t)(arg1[1]) * ((arg2[3]) * 0x2));
+  x88 = ((uint64_t)(arg1[1]) * (arg2[2]));
+  x89 = ((uint64_t)(arg1[1]) * ((arg2[1]) * 0x2));
+  x90 = ((uint64_t)(arg1[1]) * (arg2[0]));
+  x91 = ((uint64_t)(arg1[0]) * (arg2[9]));
+  x92 = ((uint64_t)(arg1[0]) * (arg2[8]));
+  x93 = ((uint64_t)(arg1[0]) * (arg2[7]));
+  x94 = ((uint64_t)(arg1[0]) * (arg2[6]));
+  x95 = ((uint64_t)(arg1[0]) * (arg2[5]));
+  x96 = ((uint64_t)(arg1[0]) * (arg2[4]));
+  x97 = ((uint64_t)(arg1[0]) * (arg2[3]));
+  x98 = ((uint64_t)(arg1[0]) * (arg2[2]));
+  x99 = ((uint64_t)(arg1[0]) * (arg2[1]));
+  x100 = ((uint64_t)(arg1[0]) * (arg2[0]));
+  x101 = (x100 + (x45 + (x44 + (x42 + (x39 + (x35 + (x30 + (x24 + (x17 + x9)))))))));
+  x102 = (x101 >> 26);
+  x103 = (uint32_t)(x101 & UINT32_C(0x3ffffff));
+  x104 = (x91 + (x82 + (x74 + (x67 + (x61 + (x56 + (x52 + (x49 + (x47 + x46)))))))));
+  x105 = (x92 + (x83 + (x75 + (x68 + (x62 + (x57 + (x53 + (x50 + (x48 + x1)))))))));
+  x106 = (x93 + (x84 + (x76 + (x69 + (x63 + (x58 + (x54 + (x51 + (x10 + x2)))))))));
+  x107 = (x94 + (x85 + (x77 + (x70 + (x64 + (x59 + (x55 + (x18 + (x11 + x3)))))))));
+  x108 = (x95 + (x86 + (x78 + (x71 + (x65 + (x60 + (x25 + (x19 + (x12 + x4)))))))));
+  x109 = (x96 + (x87 + (x79 + (x72 + (x66 + (x31 + (x26 + (x20 + (x13 + x5)))))))));
+  x110 = (x97 + (x88 + (x80 + (x73 + (x36 + (x32 + (x27 + (x21 + (x14 + x6)))))))));
+  x111 = (x98 + (x89 + (x81 + (x40 + (x37 + (x33 + (x28 + (x22 + (x15 + x7)))))))));
+  x112 = (x99 + (x90 + (x43 + (x41 + (x38 + (x34 + (x29 + (x23 + (x16 + x8)))))))));
+  x113 = (x102 + x112);
+  x114 = (x113 >> 25);
+  x115 = (uint32_t)(x113 & UINT32_C(0x1ffffff));
+  x116 = (x114 + x111);
+  x117 = (x116 >> 26);
+  x118 = (uint32_t)(x116 & UINT32_C(0x3ffffff));
+  x119 = (x117 + x110);
+  x120 = (x119 >> 25);
+  x121 = (uint32_t)(x119 & UINT32_C(0x1ffffff));
+  x122 = (x120 + x109);
+  x123 = (x122 >> 26);
+  x124 = (uint32_t)(x122 & UINT32_C(0x3ffffff));
+  x125 = (x123 + x108);
+  x126 = (x125 >> 25);
+  x127 = (uint32_t)(x125 & UINT32_C(0x1ffffff));
+  x128 = (x126 + x107);
+  x129 = (x128 >> 26);
+  x130 = (uint32_t)(x128 & UINT32_C(0x3ffffff));
+  x131 = (x129 + x106);
+  x132 = (x131 >> 25);
+  x133 = (uint32_t)(x131 & UINT32_C(0x1ffffff));
+  x134 = (x132 + x105);
+  x135 = (x134 >> 26);
+  x136 = (uint32_t)(x134 & UINT32_C(0x3ffffff));
+  x137 = (x135 + x104);
+  x138 = (x137 >> 25);
+  x139 = (uint32_t)(x137 & UINT32_C(0x1ffffff));
+  x140 = (x138 * UINT8_C(0x13));
+  x141 = (x103 + x140);
+  x142 = (uint32_t)(x141 >> 26);
+  x143 = (uint32_t)(x141 & UINT32_C(0x3ffffff));
+  x144 = (x142 + x115);
+  x145 = (fiat_25519_uint1)(x144 >> 25);
+  x146 = (x144 & UINT32_C(0x1ffffff));
+  x147 = (x145 + x118);
   out1[0] = x143;
   out1[1] = x146;
   out1[2] = x147;
@@ -752,132 +1337,253 @@ static void fiat_25519_carry_mul(uint32_t out1[10], const uint32_t arg1[10], con
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333]]
+ * The function fiat_25519_carry_square squares a field element and reduces the result.
+ *
+ * Postconditions:
+ *   eval out1 mod m = (eval arg1 * eval arg1) mod m
+ *
  */
-static void fiat_25519_carry_square(uint32_t out1[10], const uint32_t arg1[10]) {
-  uint32_t x1 = ((arg1[9]) * (uint32_t)UINT8_C(0x13));
-  uint32_t x2 = (x1 * (uint32_t)0x2);
-  uint32_t x3 = ((arg1[9]) * (uint32_t)0x2);
-  uint32_t x4 = ((arg1[8]) * (uint32_t)UINT8_C(0x13));
-  uint64_t x5 = (x4 * (uint64_t)0x2);
-  uint32_t x6 = ((arg1[8]) * (uint32_t)0x2);
-  uint32_t x7 = ((arg1[7]) * (uint32_t)UINT8_C(0x13));
-  uint32_t x8 = (x7 * (uint32_t)0x2);
-  uint32_t x9 = ((arg1[7]) * (uint32_t)0x2);
-  uint32_t x10 = ((arg1[6]) * (uint32_t)UINT8_C(0x13));
-  uint64_t x11 = (x10 * (uint64_t)0x2);
-  uint32_t x12 = ((arg1[6]) * (uint32_t)0x2);
-  uint32_t x13 = ((arg1[5]) * (uint32_t)UINT8_C(0x13));
-  uint32_t x14 = ((arg1[5]) * (uint32_t)0x2);
-  uint32_t x15 = ((arg1[4]) * (uint32_t)0x2);
-  uint32_t x16 = ((arg1[3]) * (uint32_t)0x2);
-  uint32_t x17 = ((arg1[2]) * (uint32_t)0x2);
-  uint32_t x18 = ((arg1[1]) * (uint32_t)0x2);
-  uint64_t x19 = ((uint64_t)(arg1[9]) * (x1 * (uint32_t)0x2));
-  uint64_t x20 = ((uint64_t)(arg1[8]) * x2);
-  uint64_t x21 = ((uint64_t)(arg1[8]) * x4);
-  uint64_t x22 = ((arg1[7]) * (x2 * (uint64_t)0x2));
-  uint64_t x23 = ((arg1[7]) * x5);
-  uint64_t x24 = ((uint64_t)(arg1[7]) * (x7 * (uint32_t)0x2));
-  uint64_t x25 = ((uint64_t)(arg1[6]) * x2);
-  uint64_t x26 = ((arg1[6]) * x5);
-  uint64_t x27 = ((uint64_t)(arg1[6]) * x8);
-  uint64_t x28 = ((uint64_t)(arg1[6]) * x10);
-  uint64_t x29 = ((arg1[5]) * (x2 * (uint64_t)0x2));
-  uint64_t x30 = ((arg1[5]) * x5);
-  uint64_t x31 = ((arg1[5]) * (x8 * (uint64_t)0x2));
-  uint64_t x32 = ((arg1[5]) * x11);
-  uint64_t x33 = ((uint64_t)(arg1[5]) * (x13 * (uint32_t)0x2));
-  uint64_t x34 = ((uint64_t)(arg1[4]) * x2);
-  uint64_t x35 = ((arg1[4]) * x5);
-  uint64_t x36 = ((uint64_t)(arg1[4]) * x8);
-  uint64_t x37 = ((arg1[4]) * x11);
-  uint64_t x38 = ((uint64_t)(arg1[4]) * x14);
-  uint64_t x39 = ((uint64_t)(arg1[4]) * (arg1[4]));
-  uint64_t x40 = ((arg1[3]) * (x2 * (uint64_t)0x2));
-  uint64_t x41 = ((arg1[3]) * x5);
-  uint64_t x42 = ((arg1[3]) * (x8 * (uint64_t)0x2));
-  uint64_t x43 = ((uint64_t)(arg1[3]) * x12);
-  uint64_t x44 = ((uint64_t)(arg1[3]) * (x14 * (uint32_t)0x2));
-  uint64_t x45 = ((uint64_t)(arg1[3]) * x15);
-  uint64_t x46 = ((uint64_t)(arg1[3]) * ((arg1[3]) * (uint32_t)0x2));
-  uint64_t x47 = ((uint64_t)(arg1[2]) * x2);
-  uint64_t x48 = ((arg1[2]) * x5);
-  uint64_t x49 = ((uint64_t)(arg1[2]) * x9);
-  uint64_t x50 = ((uint64_t)(arg1[2]) * x12);
-  uint64_t x51 = ((uint64_t)(arg1[2]) * x14);
-  uint64_t x52 = ((uint64_t)(arg1[2]) * x15);
-  uint64_t x53 = ((uint64_t)(arg1[2]) * x16);
-  uint64_t x54 = ((uint64_t)(arg1[2]) * (arg1[2]));
-  uint64_t x55 = ((arg1[1]) * (x2 * (uint64_t)0x2));
-  uint64_t x56 = ((uint64_t)(arg1[1]) * x6);
-  uint64_t x57 = ((uint64_t)(arg1[1]) * (x9 * (uint32_t)0x2));
-  uint64_t x58 = ((uint64_t)(arg1[1]) * x12);
-  uint64_t x59 = ((uint64_t)(arg1[1]) * (x14 * (uint32_t)0x2));
-  uint64_t x60 = ((uint64_t)(arg1[1]) * x15);
-  uint64_t x61 = ((uint64_t)(arg1[1]) * (x16 * (uint32_t)0x2));
-  uint64_t x62 = ((uint64_t)(arg1[1]) * x17);
-  uint64_t x63 = ((uint64_t)(arg1[1]) * ((arg1[1]) * (uint32_t)0x2));
-  uint64_t x64 = ((uint64_t)(arg1[0]) * x3);
-  uint64_t x65 = ((uint64_t)(arg1[0]) * x6);
-  uint64_t x66 = ((uint64_t)(arg1[0]) * x9);
-  uint64_t x67 = ((uint64_t)(arg1[0]) * x12);
-  uint64_t x68 = ((uint64_t)(arg1[0]) * x14);
-  uint64_t x69 = ((uint64_t)(arg1[0]) * x15);
-  uint64_t x70 = ((uint64_t)(arg1[0]) * x16);
-  uint64_t x71 = ((uint64_t)(arg1[0]) * x17);
-  uint64_t x72 = ((uint64_t)(arg1[0]) * x18);
-  uint64_t x73 = ((uint64_t)(arg1[0]) * (arg1[0]));
-  uint64_t x74 = (x73 + (x55 + (x48 + (x42 + (x37 + x33)))));
-  uint64_t x75 = (x74 >> 26);
-  uint32_t x76 = (uint32_t)(x74 & UINT32_C(0x3ffffff));
-  uint64_t x77 = (x64 + (x56 + (x49 + (x43 + x38))));
-  uint64_t x78 = (x65 + (x57 + (x50 + (x44 + (x39 + x19)))));
-  uint64_t x79 = (x66 + (x58 + (x51 + (x45 + x20))));
-  uint64_t x80 = (x67 + (x59 + (x52 + (x46 + (x22 + x21)))));
-  uint64_t x81 = (x68 + (x60 + (x53 + (x25 + x23))));
-  uint64_t x82 = (x69 + (x61 + (x54 + (x29 + (x26 + x24)))));
-  uint64_t x83 = (x70 + (x62 + (x34 + (x30 + x27))));
-  uint64_t x84 = (x71 + (x63 + (x40 + (x35 + (x31 + x28)))));
-  uint64_t x85 = (x72 + (x47 + (x41 + (x36 + x32))));
-  uint64_t x86 = (x75 + x85);
-  uint64_t x87 = (x86 >> 25);
-  uint32_t x88 = (uint32_t)(x86 & UINT32_C(0x1ffffff));
-  uint64_t x89 = (x87 + x84);
-  uint64_t x90 = (x89 >> 26);
-  uint32_t x91 = (uint32_t)(x89 & UINT32_C(0x3ffffff));
-  uint64_t x92 = (x90 + x83);
-  uint64_t x93 = (x92 >> 25);
-  uint32_t x94 = (uint32_t)(x92 & UINT32_C(0x1ffffff));
-  uint64_t x95 = (x93 + x82);
-  uint64_t x96 = (x95 >> 26);
-  uint32_t x97 = (uint32_t)(x95 & UINT32_C(0x3ffffff));
-  uint64_t x98 = (x96 + x81);
-  uint64_t x99 = (x98 >> 25);
-  uint32_t x100 = (uint32_t)(x98 & UINT32_C(0x1ffffff));
-  uint64_t x101 = (x99 + x80);
-  uint64_t x102 = (x101 >> 26);
-  uint32_t x103 = (uint32_t)(x101 & UINT32_C(0x3ffffff));
-  uint64_t x104 = (x102 + x79);
-  uint64_t x105 = (x104 >> 25);
-  uint32_t x106 = (uint32_t)(x104 & UINT32_C(0x1ffffff));
-  uint64_t x107 = (x105 + x78);
-  uint64_t x108 = (x107 >> 26);
-  uint32_t x109 = (uint32_t)(x107 & UINT32_C(0x3ffffff));
-  uint64_t x110 = (x108 + x77);
-  uint64_t x111 = (x110 >> 25);
-  uint32_t x112 = (uint32_t)(x110 & UINT32_C(0x1ffffff));
-  uint64_t x113 = (x111 * (uint64_t)UINT8_C(0x13));
-  uint64_t x114 = (x76 + x113);
-  uint32_t x115 = (uint32_t)(x114 >> 26);
-  uint32_t x116 = (uint32_t)(x114 & UINT32_C(0x3ffffff));
-  uint32_t x117 = (x115 + x88);
-  uint32_t x118 = (x117 >> 25);
-  uint32_t x119 = (x117 & UINT32_C(0x1ffffff));
-  uint32_t x120 = (x118 + x91);
+static FIAT_25519_FIAT_INLINE void fiat_25519_carry_square(fiat_25519_tight_field_element out1, const fiat_25519_loose_field_element arg1) {
+  uint32_t x1;
+  uint32_t x2;
+  uint32_t x3;
+  uint32_t x4;
+  uint64_t x5;
+  uint32_t x6;
+  uint32_t x7;
+  uint32_t x8;
+  uint32_t x9;
+  uint32_t x10;
+  uint64_t x11;
+  uint32_t x12;
+  uint32_t x13;
+  uint32_t x14;
+  uint32_t x15;
+  uint32_t x16;
+  uint32_t x17;
+  uint32_t x18;
+  uint64_t x19;
+  uint64_t x20;
+  uint64_t x21;
+  uint64_t x22;
+  uint64_t x23;
+  uint64_t x24;
+  uint64_t x25;
+  uint64_t x26;
+  uint64_t x27;
+  uint64_t x28;
+  uint64_t x29;
+  uint64_t x30;
+  uint64_t x31;
+  uint64_t x32;
+  uint64_t x33;
+  uint64_t x34;
+  uint64_t x35;
+  uint64_t x36;
+  uint64_t x37;
+  uint64_t x38;
+  uint64_t x39;
+  uint64_t x40;
+  uint64_t x41;
+  uint64_t x42;
+  uint64_t x43;
+  uint64_t x44;
+  uint64_t x45;
+  uint64_t x46;
+  uint64_t x47;
+  uint64_t x48;
+  uint64_t x49;
+  uint64_t x50;
+  uint64_t x51;
+  uint64_t x52;
+  uint64_t x53;
+  uint64_t x54;
+  uint64_t x55;
+  uint64_t x56;
+  uint64_t x57;
+  uint64_t x58;
+  uint64_t x59;
+  uint64_t x60;
+  uint64_t x61;
+  uint64_t x62;
+  uint64_t x63;
+  uint64_t x64;
+  uint64_t x65;
+  uint64_t x66;
+  uint64_t x67;
+  uint64_t x68;
+  uint64_t x69;
+  uint64_t x70;
+  uint64_t x71;
+  uint64_t x72;
+  uint64_t x73;
+  uint64_t x74;
+  uint64_t x75;
+  uint32_t x76;
+  uint64_t x77;
+  uint64_t x78;
+  uint64_t x79;
+  uint64_t x80;
+  uint64_t x81;
+  uint64_t x82;
+  uint64_t x83;
+  uint64_t x84;
+  uint64_t x85;
+  uint64_t x86;
+  uint64_t x87;
+  uint32_t x88;
+  uint64_t x89;
+  uint64_t x90;
+  uint32_t x91;
+  uint64_t x92;
+  uint64_t x93;
+  uint32_t x94;
+  uint64_t x95;
+  uint64_t x96;
+  uint32_t x97;
+  uint64_t x98;
+  uint64_t x99;
+  uint32_t x100;
+  uint64_t x101;
+  uint64_t x102;
+  uint32_t x103;
+  uint64_t x104;
+  uint64_t x105;
+  uint32_t x106;
+  uint64_t x107;
+  uint64_t x108;
+  uint32_t x109;
+  uint64_t x110;
+  uint64_t x111;
+  uint32_t x112;
+  uint64_t x113;
+  uint64_t x114;
+  uint32_t x115;
+  uint32_t x116;
+  uint32_t x117;
+  fiat_25519_uint1 x118;
+  uint32_t x119;
+  uint32_t x120;
+  x1 = ((arg1[9]) * UINT8_C(0x13));
+  x2 = (x1 * 0x2);
+  x3 = ((arg1[9]) * 0x2);
+  x4 = ((arg1[8]) * UINT8_C(0x13));
+  x5 = ((uint64_t)x4 * 0x2);
+  x6 = ((arg1[8]) * 0x2);
+  x7 = ((arg1[7]) * UINT8_C(0x13));
+  x8 = (x7 * 0x2);
+  x9 = ((arg1[7]) * 0x2);
+  x10 = ((arg1[6]) * UINT8_C(0x13));
+  x11 = ((uint64_t)x10 * 0x2);
+  x12 = ((arg1[6]) * 0x2);
+  x13 = ((arg1[5]) * UINT8_C(0x13));
+  x14 = ((arg1[5]) * 0x2);
+  x15 = ((arg1[4]) * 0x2);
+  x16 = ((arg1[3]) * 0x2);
+  x17 = ((arg1[2]) * 0x2);
+  x18 = ((arg1[1]) * 0x2);
+  x19 = ((uint64_t)(arg1[9]) * (x1 * 0x2));
+  x20 = ((uint64_t)(arg1[8]) * x2);
+  x21 = ((uint64_t)(arg1[8]) * x4);
+  x22 = ((arg1[7]) * ((uint64_t)x2 * 0x2));
+  x23 = ((arg1[7]) * x5);
+  x24 = ((uint64_t)(arg1[7]) * (x7 * 0x2));
+  x25 = ((uint64_t)(arg1[6]) * x2);
+  x26 = ((arg1[6]) * x5);
+  x27 = ((uint64_t)(arg1[6]) * x8);
+  x28 = ((uint64_t)(arg1[6]) * x10);
+  x29 = ((arg1[5]) * ((uint64_t)x2 * 0x2));
+  x30 = ((arg1[5]) * x5);
+  x31 = ((arg1[5]) * ((uint64_t)x8 * 0x2));
+  x32 = ((arg1[5]) * x11);
+  x33 = ((uint64_t)(arg1[5]) * (x13 * 0x2));
+  x34 = ((uint64_t)(arg1[4]) * x2);
+  x35 = ((arg1[4]) * x5);
+  x36 = ((uint64_t)(arg1[4]) * x8);
+  x37 = ((arg1[4]) * x11);
+  x38 = ((uint64_t)(arg1[4]) * x14);
+  x39 = ((uint64_t)(arg1[4]) * (arg1[4]));
+  x40 = ((arg1[3]) * ((uint64_t)x2 * 0x2));
+  x41 = ((arg1[3]) * x5);
+  x42 = ((arg1[3]) * ((uint64_t)x8 * 0x2));
+  x43 = ((uint64_t)(arg1[3]) * x12);
+  x44 = ((uint64_t)(arg1[3]) * (x14 * 0x2));
+  x45 = ((uint64_t)(arg1[3]) * x15);
+  x46 = ((uint64_t)(arg1[3]) * ((arg1[3]) * 0x2));
+  x47 = ((uint64_t)(arg1[2]) * x2);
+  x48 = ((arg1[2]) * x5);
+  x49 = ((uint64_t)(arg1[2]) * x9);
+  x50 = ((uint64_t)(arg1[2]) * x12);
+  x51 = ((uint64_t)(arg1[2]) * x14);
+  x52 = ((uint64_t)(arg1[2]) * x15);
+  x53 = ((uint64_t)(arg1[2]) * x16);
+  x54 = ((uint64_t)(arg1[2]) * (arg1[2]));
+  x55 = ((arg1[1]) * ((uint64_t)x2 * 0x2));
+  x56 = ((uint64_t)(arg1[1]) * x6);
+  x57 = ((uint64_t)(arg1[1]) * (x9 * 0x2));
+  x58 = ((uint64_t)(arg1[1]) * x12);
+  x59 = ((uint64_t)(arg1[1]) * (x14 * 0x2));
+  x60 = ((uint64_t)(arg1[1]) * x15);
+  x61 = ((uint64_t)(arg1[1]) * (x16 * 0x2));
+  x62 = ((uint64_t)(arg1[1]) * x17);
+  x63 = ((uint64_t)(arg1[1]) * ((arg1[1]) * 0x2));
+  x64 = ((uint64_t)(arg1[0]) * x3);
+  x65 = ((uint64_t)(arg1[0]) * x6);
+  x66 = ((uint64_t)(arg1[0]) * x9);
+  x67 = ((uint64_t)(arg1[0]) * x12);
+  x68 = ((uint64_t)(arg1[0]) * x14);
+  x69 = ((uint64_t)(arg1[0]) * x15);
+  x70 = ((uint64_t)(arg1[0]) * x16);
+  x71 = ((uint64_t)(arg1[0]) * x17);
+  x72 = ((uint64_t)(arg1[0]) * x18);
+  x73 = ((uint64_t)(arg1[0]) * (arg1[0]));
+  x74 = (x73 + (x55 + (x48 + (x42 + (x37 + x33)))));
+  x75 = (x74 >> 26);
+  x76 = (uint32_t)(x74 & UINT32_C(0x3ffffff));
+  x77 = (x64 + (x56 + (x49 + (x43 + x38))));
+  x78 = (x65 + (x57 + (x50 + (x44 + (x39 + x19)))));
+  x79 = (x66 + (x58 + (x51 + (x45 + x20))));
+  x80 = (x67 + (x59 + (x52 + (x46 + (x22 + x21)))));
+  x81 = (x68 + (x60 + (x53 + (x25 + x23))));
+  x82 = (x69 + (x61 + (x54 + (x29 + (x26 + x24)))));
+  x83 = (x70 + (x62 + (x34 + (x30 + x27))));
+  x84 = (x71 + (x63 + (x40 + (x35 + (x31 + x28)))));
+  x85 = (x72 + (x47 + (x41 + (x36 + x32))));
+  x86 = (x75 + x85);
+  x87 = (x86 >> 25);
+  x88 = (uint32_t)(x86 & UINT32_C(0x1ffffff));
+  x89 = (x87 + x84);
+  x90 = (x89 >> 26);
+  x91 = (uint32_t)(x89 & UINT32_C(0x3ffffff));
+  x92 = (x90 + x83);
+  x93 = (x92 >> 25);
+  x94 = (uint32_t)(x92 & UINT32_C(0x1ffffff));
+  x95 = (x93 + x82);
+  x96 = (x95 >> 26);
+  x97 = (uint32_t)(x95 & UINT32_C(0x3ffffff));
+  x98 = (x96 + x81);
+  x99 = (x98 >> 25);
+  x100 = (uint32_t)(x98 & UINT32_C(0x1ffffff));
+  x101 = (x99 + x80);
+  x102 = (x101 >> 26);
+  x103 = (uint32_t)(x101 & UINT32_C(0x3ffffff));
+  x104 = (x102 + x79);
+  x105 = (x104 >> 25);
+  x106 = (uint32_t)(x104 & UINT32_C(0x1ffffff));
+  x107 = (x105 + x78);
+  x108 = (x107 >> 26);
+  x109 = (uint32_t)(x107 & UINT32_C(0x3ffffff));
+  x110 = (x108 + x77);
+  x111 = (x110 >> 25);
+  x112 = (uint32_t)(x110 & UINT32_C(0x1ffffff));
+  x113 = (x111 * UINT8_C(0x13));
+  x114 = (x76 + x113);
+  x115 = (uint32_t)(x114 >> 26);
+  x116 = (uint32_t)(x114 & UINT32_C(0x3ffffff));
+  x117 = (x115 + x88);
+  x118 = (fiat_25519_uint1)(x117 >> 25);
+  x119 = (x117 & UINT32_C(0x1ffffff));
+  x120 = (x118 + x91);
   out1[0] = x116;
   out1[1] = x119;
   out1[2] = x120;
@@ -891,34 +1597,57 @@ static void fiat_25519_carry_square(uint32_t out1[10], const uint32_t arg1[10])
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333]]
+ * The function fiat_25519_carry reduces a field element.
+ *
+ * Postconditions:
+ *   eval out1 mod m = eval arg1 mod m
+ *
  */
-static void fiat_25519_carry(uint32_t out1[10], const uint32_t arg1[10]) {
-  uint32_t x1 = (arg1[0]);
-  uint32_t x2 = ((x1 >> 26) + (arg1[1]));
-  uint32_t x3 = ((x2 >> 25) + (arg1[2]));
-  uint32_t x4 = ((x3 >> 26) + (arg1[3]));
-  uint32_t x5 = ((x4 >> 25) + (arg1[4]));
-  uint32_t x6 = ((x5 >> 26) + (arg1[5]));
-  uint32_t x7 = ((x6 >> 25) + (arg1[6]));
-  uint32_t x8 = ((x7 >> 26) + (arg1[7]));
-  uint32_t x9 = ((x8 >> 25) + (arg1[8]));
-  uint32_t x10 = ((x9 >> 26) + (arg1[9]));
-  uint32_t x11 = ((x1 & UINT32_C(0x3ffffff)) + ((x10 >> 25) * (uint32_t)UINT8_C(0x13)));
-  uint32_t x12 = ((x11 >> 26) + (x2 & UINT32_C(0x1ffffff)));
-  uint32_t x13 = (x11 & UINT32_C(0x3ffffff));
-  uint32_t x14 = (x12 & UINT32_C(0x1ffffff));
-  uint32_t x15 = ((x12 >> 25) + (x3 & UINT32_C(0x3ffffff)));
-  uint32_t x16 = (x4 & UINT32_C(0x1ffffff));
-  uint32_t x17 = (x5 & UINT32_C(0x3ffffff));
-  uint32_t x18 = (x6 & UINT32_C(0x1ffffff));
-  uint32_t x19 = (x7 & UINT32_C(0x3ffffff));
-  uint32_t x20 = (x8 & UINT32_C(0x1ffffff));
-  uint32_t x21 = (x9 & UINT32_C(0x3ffffff));
-  uint32_t x22 = (x10 & UINT32_C(0x1ffffff));
+static FIAT_25519_FIAT_INLINE void fiat_25519_carry(fiat_25519_tight_field_element out1, const fiat_25519_loose_field_element arg1) {
+  uint32_t x1;
+  uint32_t x2;
+  uint32_t x3;
+  uint32_t x4;
+  uint32_t x5;
+  uint32_t x6;
+  uint32_t x7;
+  uint32_t x8;
+  uint32_t x9;
+  uint32_t x10;
+  uint32_t x11;
+  uint32_t x12;
+  uint32_t x13;
+  uint32_t x14;
+  uint32_t x15;
+  uint32_t x16;
+  uint32_t x17;
+  uint32_t x18;
+  uint32_t x19;
+  uint32_t x20;
+  uint32_t x21;
+  uint32_t x22;
+  x1 = (arg1[0]);
+  x2 = ((x1 >> 26) + (arg1[1]));
+  x3 = ((x2 >> 25) + (arg1[2]));
+  x4 = ((x3 >> 26) + (arg1[3]));
+  x5 = ((x4 >> 25) + (arg1[4]));
+  x6 = ((x5 >> 26) + (arg1[5]));
+  x7 = ((x6 >> 25) + (arg1[6]));
+  x8 = ((x7 >> 26) + (arg1[7]));
+  x9 = ((x8 >> 25) + (arg1[8]));
+  x10 = ((x9 >> 26) + (arg1[9]));
+  x11 = ((x1 & UINT32_C(0x3ffffff)) + ((x10 >> 25) * UINT8_C(0x13)));
+  x12 = ((fiat_25519_uint1)(x11 >> 26) + (x2 & UINT32_C(0x1ffffff)));
+  x13 = (x11 & UINT32_C(0x3ffffff));
+  x14 = (x12 & UINT32_C(0x1ffffff));
+  x15 = ((fiat_25519_uint1)(x12 >> 25) + (x3 & UINT32_C(0x3ffffff)));
+  x16 = (x4 & UINT32_C(0x1ffffff));
+  x17 = (x5 & UINT32_C(0x3ffffff));
+  x18 = (x6 & UINT32_C(0x1ffffff));
+  x19 = (x7 & UINT32_C(0x3ffffff));
+  x20 = (x8 & UINT32_C(0x1ffffff));
+  x21 = (x9 & UINT32_C(0x3ffffff));
+  x22 = (x10 & UINT32_C(0x1ffffff));
   out1[0] = x13;
   out1[1] = x14;
   out1[2] = x15;
@@ -932,23 +1661,33 @@ static void fiat_25519_carry(uint32_t out1[10], const uint32_t arg1[10]) {
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333]]
- *   arg2: [[0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999]]
+ * The function fiat_25519_add adds two field elements.
+ *
+ * Postconditions:
+ *   eval out1 mod m = (eval arg1 + eval arg2) mod m
+ *
  */
-static void fiat_25519_add(uint32_t out1[10], const uint32_t arg1[10], const uint32_t arg2[10]) {
-  uint32_t x1 = ((arg1[0]) + (arg2[0]));
-  uint32_t x2 = ((arg1[1]) + (arg2[1]));
-  uint32_t x3 = ((arg1[2]) + (arg2[2]));
-  uint32_t x4 = ((arg1[3]) + (arg2[3]));
-  uint32_t x5 = ((arg1[4]) + (arg2[4]));
-  uint32_t x6 = ((arg1[5]) + (arg2[5]));
-  uint32_t x7 = ((arg1[6]) + (arg2[6]));
-  uint32_t x8 = ((arg1[7]) + (arg2[7]));
-  uint32_t x9 = ((arg1[8]) + (arg2[8]));
-  uint32_t x10 = ((arg1[9]) + (arg2[9]));
+static FIAT_25519_FIAT_INLINE void fiat_25519_add(fiat_25519_loose_field_element out1, const fiat_25519_tight_field_element arg1, const fiat_25519_tight_field_element arg2) {
+  uint32_t x1;
+  uint32_t x2;
+  uint32_t x3;
+  uint32_t x4;
+  uint32_t x5;
+  uint32_t x6;
+  uint32_t x7;
+  uint32_t x8;
+  uint32_t x9;
+  uint32_t x10;
+  x1 = ((arg1[0]) + (arg2[0]));
+  x2 = ((arg1[1]) + (arg2[1]));
+  x3 = ((arg1[2]) + (arg2[2]));
+  x4 = ((arg1[3]) + (arg2[3]));
+  x5 = ((arg1[4]) + (arg2[4]));
+  x6 = ((arg1[5]) + (arg2[5]));
+  x7 = ((arg1[6]) + (arg2[6]));
+  x8 = ((arg1[7]) + (arg2[7]));
+  x9 = ((arg1[8]) + (arg2[8]));
+  x10 = ((arg1[9]) + (arg2[9]));
   out1[0] = x1;
   out1[1] = x2;
   out1[2] = x3;
@@ -962,23 +1701,33 @@ static void fiat_25519_add(uint32_t out1[10], const uint32_t arg1[10], const uin
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333]]
- *   arg2: [[0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999]]
+ * The function fiat_25519_sub subtracts two field elements.
+ *
+ * Postconditions:
+ *   eval out1 mod m = (eval arg1 - eval arg2) mod m
+ *
  */
-static void fiat_25519_sub(uint32_t out1[10], const uint32_t arg1[10], const uint32_t arg2[10]) {
-  uint32_t x1 = ((UINT32_C(0x7ffffda) + (arg1[0])) - (arg2[0]));
-  uint32_t x2 = ((UINT32_C(0x3fffffe) + (arg1[1])) - (arg2[1]));
-  uint32_t x3 = ((UINT32_C(0x7fffffe) + (arg1[2])) - (arg2[2]));
-  uint32_t x4 = ((UINT32_C(0x3fffffe) + (arg1[3])) - (arg2[3]));
-  uint32_t x5 = ((UINT32_C(0x7fffffe) + (arg1[4])) - (arg2[4]));
-  uint32_t x6 = ((UINT32_C(0x3fffffe) + (arg1[5])) - (arg2[5]));
-  uint32_t x7 = ((UINT32_C(0x7fffffe) + (arg1[6])) - (arg2[6]));
-  uint32_t x8 = ((UINT32_C(0x3fffffe) + (arg1[7])) - (arg2[7]));
-  uint32_t x9 = ((UINT32_C(0x7fffffe) + (arg1[8])) - (arg2[8]));
-  uint32_t x10 = ((UINT32_C(0x3fffffe) + (arg1[9])) - (arg2[9]));
+static FIAT_25519_FIAT_INLINE void fiat_25519_sub(fiat_25519_loose_field_element out1, const fiat_25519_tight_field_element arg1, const fiat_25519_tight_field_element arg2) {
+  uint32_t x1;
+  uint32_t x2;
+  uint32_t x3;
+  uint32_t x4;
+  uint32_t x5;
+  uint32_t x6;
+  uint32_t x7;
+  uint32_t x8;
+  uint32_t x9;
+  uint32_t x10;
+  x1 = ((UINT32_C(0x7ffffda) + (arg1[0])) - (arg2[0]));
+  x2 = ((UINT32_C(0x3fffffe) + (arg1[1])) - (arg2[1]));
+  x3 = ((UINT32_C(0x7fffffe) + (arg1[2])) - (arg2[2]));
+  x4 = ((UINT32_C(0x3fffffe) + (arg1[3])) - (arg2[3]));
+  x5 = ((UINT32_C(0x7fffffe) + (arg1[4])) - (arg2[4]));
+  x6 = ((UINT32_C(0x3fffffe) + (arg1[5])) - (arg2[5]));
+  x7 = ((UINT32_C(0x7fffffe) + (arg1[6])) - (arg2[6]));
+  x8 = ((UINT32_C(0x3fffffe) + (arg1[7])) - (arg2[7]));
+  x9 = ((UINT32_C(0x7fffffe) + (arg1[8])) - (arg2[8]));
+  x10 = ((UINT32_C(0x3fffffe) + (arg1[9])) - (arg2[9]));
   out1[0] = x1;
   out1[1] = x2;
   out1[2] = x3;
@@ -992,22 +1741,33 @@ static void fiat_25519_sub(uint32_t out1[10], const uint32_t arg1[10], const uin
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999], [0x0 ~> 0xd333332], [0x0 ~> 0x6999999]]
+ * The function fiat_25519_opp negates a field element.
+ *
+ * Postconditions:
+ *   eval out1 mod m = -eval arg1 mod m
+ *
  */
-static void fiat_25519_opp(uint32_t out1[10], const uint32_t arg1[10]) {
-  uint32_t x1 = (UINT32_C(0x7ffffda) - (arg1[0]));
-  uint32_t x2 = (UINT32_C(0x3fffffe) - (arg1[1]));
-  uint32_t x3 = (UINT32_C(0x7fffffe) - (arg1[2]));
-  uint32_t x4 = (UINT32_C(0x3fffffe) - (arg1[3]));
-  uint32_t x5 = (UINT32_C(0x7fffffe) - (arg1[4]));
-  uint32_t x6 = (UINT32_C(0x3fffffe) - (arg1[5]));
-  uint32_t x7 = (UINT32_C(0x7fffffe) - (arg1[6]));
-  uint32_t x8 = (UINT32_C(0x3fffffe) - (arg1[7]));
-  uint32_t x9 = (UINT32_C(0x7fffffe) - (arg1[8]));
-  uint32_t x10 = (UINT32_C(0x3fffffe) - (arg1[9]));
+static FIAT_25519_FIAT_INLINE void fiat_25519_opp(fiat_25519_loose_field_element out1, const fiat_25519_tight_field_element arg1) {
+  uint32_t x1;
+  uint32_t x2;
+  uint32_t x3;
+  uint32_t x4;
+  uint32_t x5;
+  uint32_t x6;
+  uint32_t x7;
+  uint32_t x8;
+  uint32_t x9;
+  uint32_t x10;
+  x1 = (UINT32_C(0x7ffffda) - (arg1[0]));
+  x2 = (UINT32_C(0x3fffffe) - (arg1[1]));
+  x3 = (UINT32_C(0x7fffffe) - (arg1[2]));
+  x4 = (UINT32_C(0x3fffffe) - (arg1[3]));
+  x5 = (UINT32_C(0x7fffffe) - (arg1[4]));
+  x6 = (UINT32_C(0x3fffffe) - (arg1[5]));
+  x7 = (UINT32_C(0x7fffffe) - (arg1[6]));
+  x8 = (UINT32_C(0x3fffffe) - (arg1[7]));
+  x9 = (UINT32_C(0x7fffffe) - (arg1[8]));
+  x10 = (UINT32_C(0x3fffffe) - (arg1[9]));
   out1[0] = x1;
   out1[1] = x2;
   out1[2] = x3;
@@ -1021,269 +1781,439 @@ static void fiat_25519_opp(uint32_t out1[10], const uint32_t arg1[10]) {
 }
 
 /*
- * Input Bounds:
- *   arg1: [[0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333]]
+ * The function fiat_25519_to_bytes serializes a field element to bytes in little-endian order.
+ *
+ * Postconditions:
+ *   out1 = map (λ x, ⌊((eval arg1 mod m) mod 2^(8 * (x + 1))) / 2^(8 * x)⌋) [0..31]
+ *
  * Output Bounds:
  *   out1: [[0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0x7f]]
  */
-static void fiat_25519_to_bytes(uint8_t out1[32], const uint32_t arg1[10]) {
+static FIAT_25519_FIAT_INLINE void fiat_25519_to_bytes(uint8_t out1[32], const fiat_25519_tight_field_element arg1) {
   uint32_t x1;
   fiat_25519_uint1 x2;
-  fiat_25519_subborrowx_u26(&x1, &x2, 0x0, (arg1[0]), UINT32_C(0x3ffffed));
   uint32_t x3;
   fiat_25519_uint1 x4;
-  fiat_25519_subborrowx_u25(&x3, &x4, x2, (arg1[1]), UINT32_C(0x1ffffff));
   uint32_t x5;
   fiat_25519_uint1 x6;
-  fiat_25519_subborrowx_u26(&x5, &x6, x4, (arg1[2]), UINT32_C(0x3ffffff));
   uint32_t x7;
   fiat_25519_uint1 x8;
-  fiat_25519_subborrowx_u25(&x7, &x8, x6, (arg1[3]), UINT32_C(0x1ffffff));
   uint32_t x9;
   fiat_25519_uint1 x10;
-  fiat_25519_subborrowx_u26(&x9, &x10, x8, (arg1[4]), UINT32_C(0x3ffffff));
   uint32_t x11;
   fiat_25519_uint1 x12;
-  fiat_25519_subborrowx_u25(&x11, &x12, x10, (arg1[5]), UINT32_C(0x1ffffff));
   uint32_t x13;
   fiat_25519_uint1 x14;
-  fiat_25519_subborrowx_u26(&x13, &x14, x12, (arg1[6]), UINT32_C(0x3ffffff));
   uint32_t x15;
   fiat_25519_uint1 x16;
-  fiat_25519_subborrowx_u25(&x15, &x16, x14, (arg1[7]), UINT32_C(0x1ffffff));
   uint32_t x17;
   fiat_25519_uint1 x18;
-  fiat_25519_subborrowx_u26(&x17, &x18, x16, (arg1[8]), UINT32_C(0x3ffffff));
   uint32_t x19;
   fiat_25519_uint1 x20;
-  fiat_25519_subborrowx_u25(&x19, &x20, x18, (arg1[9]), UINT32_C(0x1ffffff));
   uint32_t x21;
-  fiat_25519_cmovznz_u32(&x21, x20, 0x0, UINT32_C(0xffffffff));
   uint32_t x22;
   fiat_25519_uint1 x23;
-  fiat_25519_addcarryx_u26(&x22, &x23, 0x0, (x21 & UINT32_C(0x3ffffed)), x1);
   uint32_t x24;
   fiat_25519_uint1 x25;
-  fiat_25519_addcarryx_u25(&x24, &x25, x23, (x21 & UINT32_C(0x1ffffff)), x3);
   uint32_t x26;
   fiat_25519_uint1 x27;
-  fiat_25519_addcarryx_u26(&x26, &x27, x25, (x21 & UINT32_C(0x3ffffff)), x5);
   uint32_t x28;
   fiat_25519_uint1 x29;
-  fiat_25519_addcarryx_u25(&x28, &x29, x27, (x21 & UINT32_C(0x1ffffff)), x7);
   uint32_t x30;
   fiat_25519_uint1 x31;
-  fiat_25519_addcarryx_u26(&x30, &x31, x29, (x21 & UINT32_C(0x3ffffff)), x9);
   uint32_t x32;
   fiat_25519_uint1 x33;
-  fiat_25519_addcarryx_u25(&x32, &x33, x31, (x21 & UINT32_C(0x1ffffff)), x11);
   uint32_t x34;
   fiat_25519_uint1 x35;
-  fiat_25519_addcarryx_u26(&x34, &x35, x33, (x21 & UINT32_C(0x3ffffff)), x13);
   uint32_t x36;
   fiat_25519_uint1 x37;
-  fiat_25519_addcarryx_u25(&x36, &x37, x35, (x21 & UINT32_C(0x1ffffff)), x15);
   uint32_t x38;
   fiat_25519_uint1 x39;
-  fiat_25519_addcarryx_u26(&x38, &x39, x37, (x21 & UINT32_C(0x3ffffff)), x17);
   uint32_t x40;
   fiat_25519_uint1 x41;
-  fiat_25519_addcarryx_u25(&x40, &x41, x39, (x21 & UINT32_C(0x1ffffff)), x19);
-  uint32_t x42 = (x40 << 6);
-  uint32_t x43 = (x38 << 4);
-  uint32_t x44 = (x36 << 3);
-  uint32_t x45 = (x34 * (uint32_t)0x2);
-  uint32_t x46 = (x30 << 6);
-  uint32_t x47 = (x28 << 5);
-  uint32_t x48 = (x26 << 3);
-  uint32_t x49 = (x24 << 2);
-  uint32_t x50 = (x22 >> 8);
-  uint8_t x51 = (uint8_t)(x22 & UINT8_C(0xff));
-  uint32_t x52 = (x50 >> 8);
-  uint8_t x53 = (uint8_t)(x50 & UINT8_C(0xff));
-  uint8_t x54 = (uint8_t)(x52 >> 8);
-  uint8_t x55 = (uint8_t)(x52 & UINT8_C(0xff));
-  uint32_t x56 = (x54 + x49);
-  uint32_t x57 = (x56 >> 8);
-  uint8_t x58 = (uint8_t)(x56 & UINT8_C(0xff));
-  uint32_t x59 = (x57 >> 8);
-  uint8_t x60 = (uint8_t)(x57 & UINT8_C(0xff));
-  uint8_t x61 = (uint8_t)(x59 >> 8);
-  uint8_t x62 = (uint8_t)(x59 & UINT8_C(0xff));
-  uint32_t x63 = (x61 + x48);
-  uint32_t x64 = (x63 >> 8);
-  uint8_t x65 = (uint8_t)(x63 & UINT8_C(0xff));
-  uint32_t x66 = (x64 >> 8);
-  uint8_t x67 = (uint8_t)(x64 & UINT8_C(0xff));
-  uint8_t x68 = (uint8_t)(x66 >> 8);
-  uint8_t x69 = (uint8_t)(x66 & UINT8_C(0xff));
-  uint32_t x70 = (x68 + x47);
-  uint32_t x71 = (x70 >> 8);
-  uint8_t x72 = (uint8_t)(x70 & UINT8_C(0xff));
-  uint32_t x73 = (x71 >> 8);
-  uint8_t x74 = (uint8_t)(x71 & UINT8_C(0xff));
-  uint8_t x75 = (uint8_t)(x73 >> 8);
-  uint8_t x76 = (uint8_t)(x73 & UINT8_C(0xff));
-  uint32_t x77 = (x75 + x46);
-  uint32_t x78 = (x77 >> 8);
-  uint8_t x79 = (uint8_t)(x77 & UINT8_C(0xff));
-  uint32_t x80 = (x78 >> 8);
-  uint8_t x81 = (uint8_t)(x78 & UINT8_C(0xff));
-  uint8_t x82 = (uint8_t)(x80 >> 8);
-  uint8_t x83 = (uint8_t)(x80 & UINT8_C(0xff));
-  uint8_t x84 = (uint8_t)(x82 & UINT8_C(0xff));
-  uint32_t x85 = (x32 >> 8);
-  uint8_t x86 = (uint8_t)(x32 & UINT8_C(0xff));
-  uint32_t x87 = (x85 >> 8);
-  uint8_t x88 = (uint8_t)(x85 & UINT8_C(0xff));
-  fiat_25519_uint1 x89 = (fiat_25519_uint1)(x87 >> 8);
-  uint8_t x90 = (uint8_t)(x87 & UINT8_C(0xff));
-  uint32_t x91 = (x89 + x45);
-  uint32_t x92 = (x91 >> 8);
-  uint8_t x93 = (uint8_t)(x91 & UINT8_C(0xff));
-  uint32_t x94 = (x92 >> 8);
-  uint8_t x95 = (uint8_t)(x92 & UINT8_C(0xff));
-  uint8_t x96 = (uint8_t)(x94 >> 8);
-  uint8_t x97 = (uint8_t)(x94 & UINT8_C(0xff));
-  uint32_t x98 = (x96 + x44);
-  uint32_t x99 = (x98 >> 8);
-  uint8_t x100 = (uint8_t)(x98 & UINT8_C(0xff));
-  uint32_t x101 = (x99 >> 8);
-  uint8_t x102 = (uint8_t)(x99 & UINT8_C(0xff));
-  uint8_t x103 = (uint8_t)(x101 >> 8);
-  uint8_t x104 = (uint8_t)(x101 & UINT8_C(0xff));
-  uint32_t x105 = (x103 + x43);
-  uint32_t x106 = (x105 >> 8);
-  uint8_t x107 = (uint8_t)(x105 & UINT8_C(0xff));
-  uint32_t x108 = (x106 >> 8);
-  uint8_t x109 = (uint8_t)(x106 & UINT8_C(0xff));
-  uint8_t x110 = (uint8_t)(x108 >> 8);
-  uint8_t x111 = (uint8_t)(x108 & UINT8_C(0xff));
-  uint32_t x112 = (x110 + x42);
-  uint32_t x113 = (x112 >> 8);
-  uint8_t x114 = (uint8_t)(x112 & UINT8_C(0xff));
-  uint32_t x115 = (x113 >> 8);
-  uint8_t x116 = (uint8_t)(x113 & UINT8_C(0xff));
-  uint8_t x117 = (uint8_t)(x115 >> 8);
-  uint8_t x118 = (uint8_t)(x115 & UINT8_C(0xff));
-  out1[0] = x51;
-  out1[1] = x53;
-  out1[2] = x55;
-  out1[3] = x58;
-  out1[4] = x60;
-  out1[5] = x62;
-  out1[6] = x65;
-  out1[7] = x67;
-  out1[8] = x69;
-  out1[9] = x72;
-  out1[10] = x74;
-  out1[11] = x76;
-  out1[12] = x79;
-  out1[13] = x81;
-  out1[14] = x83;
-  out1[15] = x84;
-  out1[16] = x86;
-  out1[17] = x88;
-  out1[18] = x90;
-  out1[19] = x93;
-  out1[20] = x95;
-  out1[21] = x97;
-  out1[22] = x100;
-  out1[23] = x102;
-  out1[24] = x104;
-  out1[25] = x107;
-  out1[26] = x109;
-  out1[27] = x111;
-  out1[28] = x114;
-  out1[29] = x116;
-  out1[30] = x118;
+  uint32_t x42;
+  uint32_t x43;
+  uint32_t x44;
+  uint32_t x45;
+  uint32_t x46;
+  uint32_t x47;
+  uint32_t x48;
+  uint32_t x49;
+  uint8_t x50;
+  uint32_t x51;
+  uint8_t x52;
+  uint32_t x53;
+  uint8_t x54;
+  uint8_t x55;
+  uint32_t x56;
+  uint8_t x57;
+  uint32_t x58;
+  uint8_t x59;
+  uint32_t x60;
+  uint8_t x61;
+  uint8_t x62;
+  uint32_t x63;
+  uint8_t x64;
+  uint32_t x65;
+  uint8_t x66;
+  uint32_t x67;
+  uint8_t x68;
+  uint8_t x69;
+  uint32_t x70;
+  uint8_t x71;
+  uint32_t x72;
+  uint8_t x73;
+  uint32_t x74;
+  uint8_t x75;
+  uint8_t x76;
+  uint32_t x77;
+  uint8_t x78;
+  uint32_t x79;
+  uint8_t x80;
+  uint32_t x81;
+  uint8_t x82;
+  uint8_t x83;
+  uint8_t x84;
+  uint32_t x85;
+  uint8_t x86;
+  uint32_t x87;
+  uint8_t x88;
+  fiat_25519_uint1 x89;
+  uint32_t x90;
+  uint8_t x91;
+  uint32_t x92;
+  uint8_t x93;
+  uint32_t x94;
+  uint8_t x95;
+  uint8_t x96;
+  uint32_t x97;
+  uint8_t x98;
+  uint32_t x99;
+  uint8_t x100;
+  uint32_t x101;
+  uint8_t x102;
+  uint8_t x103;
+  uint32_t x104;
+  uint8_t x105;
+  uint32_t x106;
+  uint8_t x107;
+  uint32_t x108;
+  uint8_t x109;
+  uint8_t x110;
+  uint32_t x111;
+  uint8_t x112;
+  uint32_t x113;
+  uint8_t x114;
+  uint32_t x115;
+  uint8_t x116;
+  uint8_t x117;
+  fiat_25519_subborrowx_u26(&x1, &x2, 0x0, (arg1[0]), UINT32_C(0x3ffffed));
+  fiat_25519_subborrowx_u25(&x3, &x4, x2, (arg1[1]), UINT32_C(0x1ffffff));
+  fiat_25519_subborrowx_u26(&x5, &x6, x4, (arg1[2]), UINT32_C(0x3ffffff));
+  fiat_25519_subborrowx_u25(&x7, &x8, x6, (arg1[3]), UINT32_C(0x1ffffff));
+  fiat_25519_subborrowx_u26(&x9, &x10, x8, (arg1[4]), UINT32_C(0x3ffffff));
+  fiat_25519_subborrowx_u25(&x11, &x12, x10, (arg1[5]), UINT32_C(0x1ffffff));
+  fiat_25519_subborrowx_u26(&x13, &x14, x12, (arg1[6]), UINT32_C(0x3ffffff));
+  fiat_25519_subborrowx_u25(&x15, &x16, x14, (arg1[7]), UINT32_C(0x1ffffff));
+  fiat_25519_subborrowx_u26(&x17, &x18, x16, (arg1[8]), UINT32_C(0x3ffffff));
+  fiat_25519_subborrowx_u25(&x19, &x20, x18, (arg1[9]), UINT32_C(0x1ffffff));
+  fiat_25519_cmovznz_u32(&x21, x20, 0x0, UINT32_C(0xffffffff));
+  fiat_25519_addcarryx_u26(&x22, &x23, 0x0, x1, (x21 & UINT32_C(0x3ffffed)));
+  fiat_25519_addcarryx_u25(&x24, &x25, x23, x3, (x21 & UINT32_C(0x1ffffff)));
+  fiat_25519_addcarryx_u26(&x26, &x27, x25, x5, (x21 & UINT32_C(0x3ffffff)));
+  fiat_25519_addcarryx_u25(&x28, &x29, x27, x7, (x21 & UINT32_C(0x1ffffff)));
+  fiat_25519_addcarryx_u26(&x30, &x31, x29, x9, (x21 & UINT32_C(0x3ffffff)));
+  fiat_25519_addcarryx_u25(&x32, &x33, x31, x11, (x21 & UINT32_C(0x1ffffff)));
+  fiat_25519_addcarryx_u26(&x34, &x35, x33, x13, (x21 & UINT32_C(0x3ffffff)));
+  fiat_25519_addcarryx_u25(&x36, &x37, x35, x15, (x21 & UINT32_C(0x1ffffff)));
+  fiat_25519_addcarryx_u26(&x38, &x39, x37, x17, (x21 & UINT32_C(0x3ffffff)));
+  fiat_25519_addcarryx_u25(&x40, &x41, x39, x19, (x21 & UINT32_C(0x1ffffff)));
+  x42 = (x40 << 6);
+  x43 = (x38 << 4);
+  x44 = (x36 << 3);
+  x45 = (x34 * (uint32_t)0x2);
+  x46 = (x30 << 6);
+  x47 = (x28 << 5);
+  x48 = (x26 << 3);
+  x49 = (x24 << 2);
+  x50 = (uint8_t)(x22 & UINT8_C(0xff));
+  x51 = (x22 >> 8);
+  x52 = (uint8_t)(x51 & UINT8_C(0xff));
+  x53 = (x51 >> 8);
+  x54 = (uint8_t)(x53 & UINT8_C(0xff));
+  x55 = (uint8_t)(x53 >> 8);
+  x56 = (x49 + (uint32_t)x55);
+  x57 = (uint8_t)(x56 & UINT8_C(0xff));
+  x58 = (x56 >> 8);
+  x59 = (uint8_t)(x58 & UINT8_C(0xff));
+  x60 = (x58 >> 8);
+  x61 = (uint8_t)(x60 & UINT8_C(0xff));
+  x62 = (uint8_t)(x60 >> 8);
+  x63 = (x48 + (uint32_t)x62);
+  x64 = (uint8_t)(x63 & UINT8_C(0xff));
+  x65 = (x63 >> 8);
+  x66 = (uint8_t)(x65 & UINT8_C(0xff));
+  x67 = (x65 >> 8);
+  x68 = (uint8_t)(x67 & UINT8_C(0xff));
+  x69 = (uint8_t)(x67 >> 8);
+  x70 = (x47 + (uint32_t)x69);
+  x71 = (uint8_t)(x70 & UINT8_C(0xff));
+  x72 = (x70 >> 8);
+  x73 = (uint8_t)(x72 & UINT8_C(0xff));
+  x74 = (x72 >> 8);
+  x75 = (uint8_t)(x74 & UINT8_C(0xff));
+  x76 = (uint8_t)(x74 >> 8);
+  x77 = (x46 + (uint32_t)x76);
+  x78 = (uint8_t)(x77 & UINT8_C(0xff));
+  x79 = (x77 >> 8);
+  x80 = (uint8_t)(x79 & UINT8_C(0xff));
+  x81 = (x79 >> 8);
+  x82 = (uint8_t)(x81 & UINT8_C(0xff));
+  x83 = (uint8_t)(x81 >> 8);
+  x84 = (uint8_t)(x32 & UINT8_C(0xff));
+  x85 = (x32 >> 8);
+  x86 = (uint8_t)(x85 & UINT8_C(0xff));
+  x87 = (x85 >> 8);
+  x88 = (uint8_t)(x87 & UINT8_C(0xff));
+  x89 = (fiat_25519_uint1)(x87 >> 8);
+  x90 = (x45 + (uint32_t)x89);
+  x91 = (uint8_t)(x90 & UINT8_C(0xff));
+  x92 = (x90 >> 8);
+  x93 = (uint8_t)(x92 & UINT8_C(0xff));
+  x94 = (x92 >> 8);
+  x95 = (uint8_t)(x94 & UINT8_C(0xff));
+  x96 = (uint8_t)(x94 >> 8);
+  x97 = (x44 + (uint32_t)x96);
+  x98 = (uint8_t)(x97 & UINT8_C(0xff));
+  x99 = (x97 >> 8);
+  x100 = (uint8_t)(x99 & UINT8_C(0xff));
+  x101 = (x99 >> 8);
+  x102 = (uint8_t)(x101 & UINT8_C(0xff));
+  x103 = (uint8_t)(x101 >> 8);
+  x104 = (x43 + (uint32_t)x103);
+  x105 = (uint8_t)(x104 & UINT8_C(0xff));
+  x106 = (x104 >> 8);
+  x107 = (uint8_t)(x106 & UINT8_C(0xff));
+  x108 = (x106 >> 8);
+  x109 = (uint8_t)(x108 & UINT8_C(0xff));
+  x110 = (uint8_t)(x108 >> 8);
+  x111 = (x42 + (uint32_t)x110);
+  x112 = (uint8_t)(x111 & UINT8_C(0xff));
+  x113 = (x111 >> 8);
+  x114 = (uint8_t)(x113 & UINT8_C(0xff));
+  x115 = (x113 >> 8);
+  x116 = (uint8_t)(x115 & UINT8_C(0xff));
+  x117 = (uint8_t)(x115 >> 8);
+  out1[0] = x50;
+  out1[1] = x52;
+  out1[2] = x54;
+  out1[3] = x57;
+  out1[4] = x59;
+  out1[5] = x61;
+  out1[6] = x64;
+  out1[7] = x66;
+  out1[8] = x68;
+  out1[9] = x71;
+  out1[10] = x73;
+  out1[11] = x75;
+  out1[12] = x78;
+  out1[13] = x80;
+  out1[14] = x82;
+  out1[15] = x83;
+  out1[16] = x84;
+  out1[17] = x86;
+  out1[18] = x88;
+  out1[19] = x91;
+  out1[20] = x93;
+  out1[21] = x95;
+  out1[22] = x98;
+  out1[23] = x100;
+  out1[24] = x102;
+  out1[25] = x105;
+  out1[26] = x107;
+  out1[27] = x109;
+  out1[28] = x112;
+  out1[29] = x114;
+  out1[30] = x116;
   out1[31] = x117;
 }
 
 /*
+ * The function fiat_25519_from_bytes deserializes a field element from bytes in little-endian order.
+ *
+ * Postconditions:
+ *   eval out1 mod m = bytes_eval arg1 mod m
+ *
  * Input Bounds:
  *   arg1: [[0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0xff], [0x0 ~> 0x7f]]
- * Output Bounds:
- *   out1: [[0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333], [0x0 ~> 0x4666666], [0x0 ~> 0x2333333]]
  */
-static void fiat_25519_from_bytes(uint32_t out1[10], const uint8_t arg1[32]) {
-  uint32_t x1 = ((uint32_t)(arg1[31]) << 18);
-  uint32_t x2 = ((uint32_t)(arg1[30]) << 10);
-  uint32_t x3 = ((uint32_t)(arg1[29]) << 2);
-  uint32_t x4 = ((uint32_t)(arg1[28]) << 20);
-  uint32_t x5 = ((uint32_t)(arg1[27]) << 12);
-  uint32_t x6 = ((uint32_t)(arg1[26]) << 4);
-  uint32_t x7 = ((uint32_t)(arg1[25]) << 21);
-  uint32_t x8 = ((uint32_t)(arg1[24]) << 13);
-  uint32_t x9 = ((uint32_t)(arg1[23]) << 5);
-  uint32_t x10 = ((uint32_t)(arg1[22]) << 23);
-  uint32_t x11 = ((uint32_t)(arg1[21]) << 15);
-  uint32_t x12 = ((uint32_t)(arg1[20]) << 7);
-  uint32_t x13 = ((uint32_t)(arg1[19]) << 24);
-  uint32_t x14 = ((uint32_t)(arg1[18]) << 16);
-  uint32_t x15 = ((uint32_t)(arg1[17]) << 8);
-  uint8_t x16 = (arg1[16]);
-  uint32_t x17 = ((uint32_t)(arg1[15]) << 18);
-  uint32_t x18 = ((uint32_t)(arg1[14]) << 10);
-  uint32_t x19 = ((uint32_t)(arg1[13]) << 2);
-  uint32_t x20 = ((uint32_t)(arg1[12]) << 19);
-  uint32_t x21 = ((uint32_t)(arg1[11]) << 11);
-  uint32_t x22 = ((uint32_t)(arg1[10]) << 3);
-  uint32_t x23 = ((uint32_t)(arg1[9]) << 21);
-  uint32_t x24 = ((uint32_t)(arg1[8]) << 13);
-  uint32_t x25 = ((uint32_t)(arg1[7]) << 5);
-  uint32_t x26 = ((uint32_t)(arg1[6]) << 22);
-  uint32_t x27 = ((uint32_t)(arg1[5]) << 14);
-  uint32_t x28 = ((uint32_t)(arg1[4]) << 6);
-  uint32_t x29 = ((uint32_t)(arg1[3]) << 24);
-  uint32_t x30 = ((uint32_t)(arg1[2]) << 16);
-  uint32_t x31 = ((uint32_t)(arg1[1]) << 8);
-  uint8_t x32 = (arg1[0]);
-  uint32_t x33 = (x32 + (x31 + (x30 + x29)));
-  uint8_t x34 = (uint8_t)(x33 >> 26);
-  uint32_t x35 = (x33 & UINT32_C(0x3ffffff));
-  uint32_t x36 = (x3 + (x2 + x1));
-  uint32_t x37 = (x6 + (x5 + x4));
-  uint32_t x38 = (x9 + (x8 + x7));
-  uint32_t x39 = (x12 + (x11 + x10));
-  uint32_t x40 = (x16 + (x15 + (x14 + x13)));
-  uint32_t x41 = (x19 + (x18 + x17));
-  uint32_t x42 = (x22 + (x21 + x20));
-  uint32_t x43 = (x25 + (x24 + x23));
-  uint32_t x44 = (x28 + (x27 + x26));
-  uint32_t x45 = (x34 + x44);
-  uint8_t x46 = (uint8_t)(x45 >> 25);
-  uint32_t x47 = (x45 & UINT32_C(0x1ffffff));
-  uint32_t x48 = (x46 + x43);
-  uint8_t x49 = (uint8_t)(x48 >> 26);
-  uint32_t x50 = (x48 & UINT32_C(0x3ffffff));
-  uint32_t x51 = (x49 + x42);
-  uint8_t x52 = (uint8_t)(x51 >> 25);
-  uint32_t x53 = (x51 & UINT32_C(0x1ffffff));
-  uint32_t x54 = (x52 + x41);
-  uint32_t x55 = (x54 & UINT32_C(0x3ffffff));
-  uint8_t x56 = (uint8_t)(x40 >> 25);
-  uint32_t x57 = (x40 & UINT32_C(0x1ffffff));
-  uint32_t x58 = (x56 + x39);
-  uint8_t x59 = (uint8_t)(x58 >> 26);
-  uint32_t x60 = (x58 & UINT32_C(0x3ffffff));
-  uint32_t x61 = (x59 + x38);
-  uint8_t x62 = (uint8_t)(x61 >> 25);
-  uint32_t x63 = (x61 & UINT32_C(0x1ffffff));
-  uint32_t x64 = (x62 + x37);
-  uint8_t x65 = (uint8_t)(x64 >> 26);
-  uint32_t x66 = (x64 & UINT32_C(0x3ffffff));
-  uint32_t x67 = (x65 + x36);
-  out1[0] = x35;
-  out1[1] = x47;
-  out1[2] = x50;
-  out1[3] = x53;
+static FIAT_25519_FIAT_INLINE void fiat_25519_from_bytes(fiat_25519_tight_field_element out1, const uint8_t arg1[32]) {
+  uint32_t x1;
+  uint32_t x2;
+  uint32_t x3;
+  uint32_t x4;
+  uint32_t x5;
+  uint32_t x6;
+  uint32_t x7;
+  uint32_t x8;
+  uint32_t x9;
+  uint32_t x10;
+  uint32_t x11;
+  uint32_t x12;
+  uint32_t x13;
+  uint32_t x14;
+  uint32_t x15;
+  uint8_t x16;
+  uint32_t x17;
+  uint32_t x18;
+  uint32_t x19;
+  uint32_t x20;
+  uint32_t x21;
+  uint32_t x22;
+  uint32_t x23;
+  uint32_t x24;
+  uint32_t x25;
+  uint32_t x26;
+  uint32_t x27;
+  uint32_t x28;
+  uint32_t x29;
+  uint32_t x30;
+  uint32_t x31;
+  uint8_t x32;
+  uint32_t x33;
+  uint32_t x34;
+  uint32_t x35;
+  uint32_t x36;
+  uint8_t x37;
+  uint32_t x38;
+  uint32_t x39;
+  uint32_t x40;
+  uint32_t x41;
+  uint8_t x42;
+  uint32_t x43;
+  uint32_t x44;
+  uint32_t x45;
+  uint32_t x46;
+  uint8_t x47;
+  uint32_t x48;
+  uint32_t x49;
+  uint32_t x50;
+  uint32_t x51;
+  uint8_t x52;
+  uint32_t x53;
+  uint32_t x54;
+  uint32_t x55;
+  uint32_t x56;
+  uint32_t x57;
+  uint32_t x58;
+  uint32_t x59;
+  uint8_t x60;
+  uint32_t x61;
+  uint32_t x62;
+  uint32_t x63;
+  uint32_t x64;
+  uint8_t x65;
+  uint32_t x66;
+  uint32_t x67;
+  uint32_t x68;
+  uint32_t x69;
+  uint8_t x70;
+  uint32_t x71;
+  uint32_t x72;
+  uint32_t x73;
+  uint32_t x74;
+  uint8_t x75;
+  uint32_t x76;
+  uint32_t x77;
+  uint32_t x78;
+  x1 = ((uint32_t)(arg1[31]) << 18);
+  x2 = ((uint32_t)(arg1[30]) << 10);
+  x3 = ((uint32_t)(arg1[29]) << 2);
+  x4 = ((uint32_t)(arg1[28]) << 20);
+  x5 = ((uint32_t)(arg1[27]) << 12);
+  x6 = ((uint32_t)(arg1[26]) << 4);
+  x7 = ((uint32_t)(arg1[25]) << 21);
+  x8 = ((uint32_t)(arg1[24]) << 13);
+  x9 = ((uint32_t)(arg1[23]) << 5);
+  x10 = ((uint32_t)(arg1[22]) << 23);
+  x11 = ((uint32_t)(arg1[21]) << 15);
+  x12 = ((uint32_t)(arg1[20]) << 7);
+  x13 = ((uint32_t)(arg1[19]) << 24);
+  x14 = ((uint32_t)(arg1[18]) << 16);
+  x15 = ((uint32_t)(arg1[17]) << 8);
+  x16 = (arg1[16]);
+  x17 = ((uint32_t)(arg1[15]) << 18);
+  x18 = ((uint32_t)(arg1[14]) << 10);
+  x19 = ((uint32_t)(arg1[13]) << 2);
+  x20 = ((uint32_t)(arg1[12]) << 19);
+  x21 = ((uint32_t)(arg1[11]) << 11);
+  x22 = ((uint32_t)(arg1[10]) << 3);
+  x23 = ((uint32_t)(arg1[9]) << 21);
+  x24 = ((uint32_t)(arg1[8]) << 13);
+  x25 = ((uint32_t)(arg1[7]) << 5);
+  x26 = ((uint32_t)(arg1[6]) << 22);
+  x27 = ((uint32_t)(arg1[5]) << 14);
+  x28 = ((uint32_t)(arg1[4]) << 6);
+  x29 = ((uint32_t)(arg1[3]) << 24);
+  x30 = ((uint32_t)(arg1[2]) << 16);
+  x31 = ((uint32_t)(arg1[1]) << 8);
+  x32 = (arg1[0]);
+  x33 = (x31 + (uint32_t)x32);
+  x34 = (x30 + x33);
+  x35 = (x29 + x34);
+  x36 = (x35 & UINT32_C(0x3ffffff));
+  x37 = (uint8_t)(x35 >> 26);
+  x38 = (x28 + (uint32_t)x37);
+  x39 = (x27 + x38);
+  x40 = (x26 + x39);
+  x41 = (x40 & UINT32_C(0x1ffffff));
+  x42 = (uint8_t)(x40 >> 25);
+  x43 = (x25 + (uint32_t)x42);
+  x44 = (x24 + x43);
+  x45 = (x23 + x44);
+  x46 = (x45 & UINT32_C(0x3ffffff));
+  x47 = (uint8_t)(x45 >> 26);
+  x48 = (x22 + (uint32_t)x47);
+  x49 = (x21 + x48);
+  x50 = (x20 + x49);
+  x51 = (x50 & UINT32_C(0x1ffffff));
+  x52 = (uint8_t)(x50 >> 25);
+  x53 = (x19 + (uint32_t)x52);
+  x54 = (x18 + x53);
+  x55 = (x17 + x54);
+  x56 = (x15 + (uint32_t)x16);
+  x57 = (x14 + x56);
+  x58 = (x13 + x57);
+  x59 = (x58 & UINT32_C(0x1ffffff));
+  x60 = (uint8_t)(x58 >> 25);
+  x61 = (x12 + (uint32_t)x60);
+  x62 = (x11 + x61);
+  x63 = (x10 + x62);
+  x64 = (x63 & UINT32_C(0x3ffffff));
+  x65 = (uint8_t)(x63 >> 26);
+  x66 = (x9 + (uint32_t)x65);
+  x67 = (x8 + x66);
+  x68 = (x7 + x67);
+  x69 = (x68 & UINT32_C(0x1ffffff));
+  x70 = (uint8_t)(x68 >> 25);
+  x71 = (x6 + (uint32_t)x70);
+  x72 = (x5 + x71);
+  x73 = (x4 + x72);
+  x74 = (x73 & UINT32_C(0x3ffffff));
+  x75 = (uint8_t)(x73 >> 26);
+  x76 = (x3 + (uint32_t)x75);
+  x77 = (x2 + x76);
+  x78 = (x1 + x77);
+  out1[0] = x36;
+  out1[1] = x41;
+  out1[2] = x46;
+  out1[3] = x51;
   out1[4] = x55;
-  out1[5] = x57;
-  out1[6] = x60;
-  out1[7] = x63;
-  out1[8] = x66;
-  out1[9] = x67;
+  out1[5] = x59;
+  out1[6] = x64;
+  out1[7] = x69;
+  out1[8] = x74;
+  out1[9] = x78;
 }
 
 #endif /* not defined(BORINGSSL_CURVE25519_64BIT) */
diff --git a/src/plugins/preauth/spake/edwards25519_tables.h b/src/plugins/preauth/spake/edwards25519_tables.h
index c6c501373..20f0fba4c 100644
--- a/src/plugins/preauth/spake/edwards25519_tables.h
+++ b/src/plugins/preauth/spake/edwards25519_tables.h
@@ -23,7 +23,7 @@
  * SOFTWARE.
  */
 
-/* From BoringSSL third-party/fiat/curve25519_tables.h */
+/* From BoringSSL crypto/curve25519/curve25519_tables.h */
 
 static const fe d = {{
 #if defined(BORINGSSL_CURVE25519_64BIT)


More information about the cvs-krb5 mailing list