Add a rewrite rule to collapse constant casts

If, e.g., we know from bounds analysis that the result of an operation
fits in the range r[0~>0], we now just replace it with the literal
constant.

Fixes #493

After     | File Name                                    | Before    || Change    | % Change
--------------------------------------------------------------------------------------------
21m22.14s | Total                                        | 21m22.79s || -0m00.65s | -0.05%
--------------------------------------------------------------------------------------------
4m09.97s  | PushButtonSynthesis.vo                       | 4m10.56s  || -0m00.59s | -0.23%
3m09.12s  | p384_32.c                                    | 3m08.91s  || +0m00.21s | +0.11%
2m05.94s  | Rewriter.vo                                  | 2m06.30s  || -0m00.35s | -0.28%
1m56.58s  | RewriterWf2.vo                               | 1m56.09s  || +0m00.48s | +0.42%
1m52.39s  | RewriterRulesGood.vo                         | 1m52.04s  || +0m00.35s | +0.31%
1m46.01s  | RewriterRulesInterpGood.vo                   | 1m45.79s  || +0m00.21s | +0.20%
0m46.44s  | RewriterInterpProofs1.vo                     | 0m46.47s  || -0m00.03s | -0.06%
0m44.96s  | ExtractionHaskell/word_by_word_montgomery    | 0m45.59s  || -0m00.63s | -1.38%
0m39.18s  | p521_32.c                                    | 0m39.33s  || -0m00.14s | -0.38%
0m32.41s  | p521_64.c                                    | 0m32.54s  || -0m00.13s | -0.39%
0m30.87s  | ExtractionHaskell/unsaturated_solinas        | 0m30.67s  || +0m00.19s | +0.65%
0m24.32s  | ExtractionHaskell/saturated_solinas          | 0m24.44s  || -0m00.12s | -0.49%
0m23.59s  | RewriterWf1.vo                               | 0m24.10s  || -0m00.51s | -2.11%
0m17.01s  | ExtractionOCaml/word_by_word_montgomery      | 0m17.14s  || -0m00.12s | -0.75%
0m13.48s  | secp256k1_32.c                               | 0m13.30s  || +0m00.17s | +1.35%
0m13.11s  | p256_32.c                                    | 0m13.37s  || -0m00.25s | -1.94%
0m11.34s  | p484_64.c                                    | 0m11.34s  || +0m00.00s | +0.00%
0m10.78s  | ExtractionOCaml/unsaturated_solinas          | 0m10.79s  || -0m00.00s | -0.09%
0m10.27s  | ExtractionOCaml/word_by_word_montgomery.ml   | 0m10.06s  || +0m00.20s | +2.08%
0m08.11s  | ExtractionOCaml/saturated_solinas            | 0m07.92s  || +0m00.18s | +2.39%
0m06.92s  | ExtractionOCaml/unsaturated_solinas.ml       | 0m07.02s  || -0m00.09s | -1.42%
0m06.18s  | ExtractionHaskell/word_by_word_montgomery.hs | 0m06.28s  || -0m00.10s | -1.59%
0m06.13s  | BoundsPipeline.vo                            | 0m05.98s  || +0m00.14s | +2.50%
0m05.90s  | p224_32.c                                    | 0m05.92s  || -0m00.01s | -0.33%
0m05.29s  | p384_64.c                                    | 0m05.33s  || -0m00.04s | -0.75%
0m05.17s  | ExtractionOCaml/saturated_solinas.ml         | 0m05.20s  || -0m00.03s | -0.57%
0m04.91s  | ExtractionHaskell/unsaturated_solinas.hs     | 0m04.93s  || -0m00.01s | -0.40%
0m04.06s  | ExtractionHaskell/saturated_solinas.hs       | 0m04.00s  || +0m00.05s | +1.49%
0m02.21s  | curve25519_32.c                              | 0m02.22s  || -0m00.01s | -0.45%
0m01.52s  | curve25519_64.c                              | 0m01.50s  || +0m00.02s | +1.33%
0m01.38s  | CLI.vo                                       | 0m01.42s  || -0m00.04s | -2.81%
0m01.14s  | RewriterProofs.vo                            | 0m01.13s  || +0m00.01s | +0.88%
0m01.14s  | StandaloneOCamlMain.vo                       | 0m00.96s  || +0m00.17s | +18.74%
0m01.12s  | StandaloneHaskellMain.vo                     | 0m01.03s  || +0m00.09s | +8.73%
0m01.12s  | secp256k1_64.c                               | 0m01.00s  || +0m00.12s | +12.00%
0m01.05s  | p256_64.c                                    | 0m00.98s  || +0m00.07s | +7.14%
0m01.03s  | p224_64.c                                    | 0m01.15s  || -0m00.11s | -10.43%

1 files changed, 125 insertions, 129 deletions

diff --git a/p521_64.c b/p521_64.c
index f81fa7967..08ba9e1b7 100644
--- a/p521_64.c
+++ b/p521_64.c
@@ -646,73 +646,73 @@ static void fiat_p521_to_bytes(uint8_t out1[66], const uint64_t arg1[9]) {
   uint8_t x100 = (uint8_t)(x97 & UINT8_C(0xff));
   uint8_t x101 = (uint8_t)(x99 >> 8);
   uint8_t x102 = (uint8_t)(x99 & UINT8_C(0xff));
-  fiat_p521_uint1 x103 = (fiat_p521_uint1)(x101 >> 8);
-  uint8_t x104 = (uint8_t)(x101 & UINT8_C(0xff));
-  uint64_t x105 = (x103 + x28);
-  uint64_t x106 = (x105 >> 8);
-  uint8_t x107 = (uint8_t)(x105 & UINT8_C(0xff));
-  uint64_t x108 = (x106 >> 8);
-  uint8_t x109 = (uint8_t)(x106 & UINT8_C(0xff));
-  uint64_t x110 = (x108 >> 8);
-  uint8_t x111 = (uint8_t)(x108 & UINT8_C(0xff));
-  uint64_t x112 = (x110 >> 8);
-  uint8_t x113 = (uint8_t)(x110 & UINT8_C(0xff));
-  uint64_t x114 = (x112 >> 8);
-  uint8_t x115 = (uint8_t)(x112 & UINT8_C(0xff));
-  uint64_t x116 = (x114 >> 8);
-  uint8_t x117 = (uint8_t)(x114 & UINT8_C(0xff));
-  uint8_t x118 = (uint8_t)(x116 >> 8);
-  uint8_t x119 = (uint8_t)(x116 & UINT8_C(0xff));
-  uint64_t x120 = (x118 + x40);
-  uint64_t x121 = (x120 >> 8);
-  uint8_t x122 = (uint8_t)(x120 & UINT8_C(0xff));
-  uint64_t x123 = (x121 >> 8);
-  uint8_t x124 = (uint8_t)(x121 & UINT8_C(0xff));
-  uint64_t x125 = (x123 >> 8);
-  uint8_t x126 = (uint8_t)(x123 & UINT8_C(0xff));
-  uint64_t x127 = (x125 >> 8);
-  uint8_t x128 = (uint8_t)(x125 & UINT8_C(0xff));
-  uint64_t x129 = (x127 >> 8);
-  uint8_t x130 = (uint8_t)(x127 & UINT8_C(0xff));
-  uint64_t x131 = (x129 >> 8);
-  uint8_t x132 = (uint8_t)(x129 & UINT8_C(0xff));
-  uint8_t x133 = (uint8_t)(x131 >> 8);
-  uint8_t x134 = (uint8_t)(x131 & UINT8_C(0xff));
-  uint64_t x135 = (x133 + x39);
-  uint64_t x136 = (x135 >> 8);
-  uint8_t x137 = (uint8_t)(x135 & UINT8_C(0xff));
-  uint64_t x138 = (x136 >> 8);
-  uint8_t x139 = (uint8_t)(x136 & UINT8_C(0xff));
-  uint64_t x140 = (x138 >> 8);
-  uint8_t x141 = (uint8_t)(x138 & UINT8_C(0xff));
-  uint64_t x142 = (x140 >> 8);
-  uint8_t x143 = (uint8_t)(x140 & UINT8_C(0xff));
-  uint64_t x144 = (x142 >> 8);
-  uint8_t x145 = (uint8_t)(x142 & UINT8_C(0xff));
-  uint64_t x146 = (x144 >> 8);
-  uint8_t x147 = (uint8_t)(x144 & UINT8_C(0xff));
-  uint8_t x148 = (uint8_t)(x146 >> 8);
-  uint8_t x149 = (uint8_t)(x146 & UINT8_C(0xff));
-  uint64_t x150 = (x148 + x38);
-  uint64_t x151 = (x150 >> 8);
-  uint8_t x152 = (uint8_t)(x150 & UINT8_C(0xff));
-  uint64_t x153 = (x151 >> 8);
-  uint8_t x154 = (uint8_t)(x151 & UINT8_C(0xff));
-  uint64_t x155 = (x153 >> 8);
-  uint8_t x156 = (uint8_t)(x153 & UINT8_C(0xff));
-  uint64_t x157 = (x155 >> 8);
-  uint8_t x158 = (uint8_t)(x155 & UINT8_C(0xff));
-  uint64_t x159 = (x157 >> 8);
-  uint8_t x160 = (uint8_t)(x157 & UINT8_C(0xff));
-  uint64_t x161 = (x159 >> 8);
-  uint8_t x162 = (uint8_t)(x159 & UINT8_C(0xff));
-  uint8_t x163 = (uint8_t)(x161 >> 8);
-  uint8_t x164 = (uint8_t)(x161 & UINT8_C(0xff));
-  fiat_p521_uint1 x165 = (fiat_p521_uint1)(x163 >> 8);
-  uint8_t x166 = (uint8_t)(x163 & UINT8_C(0xff));
-  uint64_t x167 = (x165 + x36);
-  uint64_t x168 = (x167 >> 8);
-  uint8_t x169 = (uint8_t)(x167 & UINT8_C(0xff));
+  uint8_t x103 = (uint8_t)(x101 & UINT8_C(0xff));
+  uint64_t x104 = (0x0 + x28);
+  uint64_t x105 = (x104 >> 8);
+  uint8_t x106 = (uint8_t)(x104 & UINT8_C(0xff));
+  uint64_t x107 = (x105 >> 8);
+  uint8_t x108 = (uint8_t)(x105 & UINT8_C(0xff));
+  uint64_t x109 = (x107 >> 8);
+  uint8_t x110 = (uint8_t)(x107 & UINT8_C(0xff));
+  uint64_t x111 = (x109 >> 8);
+  uint8_t x112 = (uint8_t)(x109 & UINT8_C(0xff));
+  uint64_t x113 = (x111 >> 8);
+  uint8_t x114 = (uint8_t)(x111 & UINT8_C(0xff));
+  uint64_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));
+  uint64_t x119 = (x117 + x40);
+  uint64_t x120 = (x119 >> 8);
+  uint8_t x121 = (uint8_t)(x119 & UINT8_C(0xff));
+  uint64_t x122 = (x120 >> 8);
+  uint8_t x123 = (uint8_t)(x120 & UINT8_C(0xff));
+  uint64_t x124 = (x122 >> 8);
+  uint8_t x125 = (uint8_t)(x122 & UINT8_C(0xff));
+  uint64_t x126 = (x124 >> 8);
+  uint8_t x127 = (uint8_t)(x124 & UINT8_C(0xff));
+  uint64_t x128 = (x126 >> 8);
+  uint8_t x129 = (uint8_t)(x126 & UINT8_C(0xff));
+  uint64_t x130 = (x128 >> 8);
+  uint8_t x131 = (uint8_t)(x128 & UINT8_C(0xff));
+  uint8_t x132 = (uint8_t)(x130 >> 8);
+  uint8_t x133 = (uint8_t)(x130 & UINT8_C(0xff));
+  uint64_t x134 = (x132 + x39);
+  uint64_t x135 = (x134 >> 8);
+  uint8_t x136 = (uint8_t)(x134 & UINT8_C(0xff));
+  uint64_t x137 = (x135 >> 8);
+  uint8_t x138 = (uint8_t)(x135 & UINT8_C(0xff));
+  uint64_t x139 = (x137 >> 8);
+  uint8_t x140 = (uint8_t)(x137 & UINT8_C(0xff));
+  uint64_t x141 = (x139 >> 8);
+  uint8_t x142 = (uint8_t)(x139 & UINT8_C(0xff));
+  uint64_t x143 = (x141 >> 8);
+  uint8_t x144 = (uint8_t)(x141 & UINT8_C(0xff));
+  uint64_t x145 = (x143 >> 8);
+  uint8_t x146 = (uint8_t)(x143 & UINT8_C(0xff));
+  uint8_t x147 = (uint8_t)(x145 >> 8);
+  uint8_t x148 = (uint8_t)(x145 & UINT8_C(0xff));
+  uint64_t x149 = (x147 + x38);
+  uint64_t x150 = (x149 >> 8);
+  uint8_t x151 = (uint8_t)(x149 & UINT8_C(0xff));
+  uint64_t x152 = (x150 >> 8);
+  uint8_t x153 = (uint8_t)(x150 & UINT8_C(0xff));
+  uint64_t x154 = (x152 >> 8);
+  uint8_t x155 = (uint8_t)(x152 & UINT8_C(0xff));
+  uint64_t x156 = (x154 >> 8);
+  uint8_t x157 = (uint8_t)(x154 & UINT8_C(0xff));
+  uint64_t x158 = (x156 >> 8);
+  uint8_t x159 = (uint8_t)(x156 & UINT8_C(0xff));
+  uint64_t x160 = (x158 >> 8);
+  uint8_t x161 = (uint8_t)(x158 & UINT8_C(0xff));
+  uint8_t x162 = (uint8_t)(x160 >> 8);
+  uint8_t x163 = (uint8_t)(x160 & UINT8_C(0xff));
+  uint8_t x164 = (uint8_t)(x162 & UINT8_C(0xff));
+  uint64_t x165 = (0x0 + x36);
+  uint64_t x166 = (x165 >> 8);
+  uint8_t x167 = (uint8_t)(x165 & UINT8_C(0xff));
+  uint64_t x168 = (x166 >> 8);
+  uint8_t x169 = (uint8_t)(x166 & UINT8_C(0xff));
   uint64_t x170 = (x168 >> 8);
   uint8_t x171 = (uint8_t)(x168 & UINT8_C(0xff));
   uint64_t x172 = (x170 >> 8);
@@ -721,10 +721,8 @@ static void fiat_p521_to_bytes(uint8_t out1[66], const uint64_t arg1[9]) {
   uint8_t x175 = (uint8_t)(x172 & UINT8_C(0xff));
   uint64_t x176 = (x174 >> 8);
   uint8_t x177 = (uint8_t)(x174 & UINT8_C(0xff));
-  uint64_t x178 = (x176 >> 8);
+  fiat_p521_uint1 x178 = (fiat_p521_uint1)(x176 >> 8);
   uint8_t x179 = (uint8_t)(x176 & UINT8_C(0xff));
-  fiat_p521_uint1 x180 = (fiat_p521_uint1)(x178 >> 8);
-  uint8_t x181 = (uint8_t)(x178 & UINT8_C(0xff));
   out1[0] = x45;
   out1[1] = x47;
   out1[2] = x49;
@@ -753,44 +751,44 @@ static void fiat_p521_to_bytes(uint8_t out1[66], const uint64_t arg1[9]) {
   out1[25] = x98;
   out1[26] = x100;
   out1[27] = x102;
-  out1[28] = x104;
-  out1[29] = x107;
-  out1[30] = x109;
-  out1[31] = x111;
-  out1[32] = x113;
-  out1[33] = x115;
-  out1[34] = x117;
-  out1[35] = x119;
-  out1[36] = x122;
-  out1[37] = x124;
-  out1[38] = x126;
-  out1[39] = x128;
-  out1[40] = x130;
-  out1[41] = x132;
-  out1[42] = x134;
-  out1[43] = x137;
-  out1[44] = x139;
-  out1[45] = x141;
-  out1[46] = x143;
-  out1[47] = x145;
-  out1[48] = x147;
-  out1[49] = x149;
-  out1[50] = x152;
-  out1[51] = x154;
-  out1[52] = x156;
-  out1[53] = x158;
-  out1[54] = x160;
-  out1[55] = x162;
-  out1[56] = x164;
-  out1[57] = x166;
-  out1[58] = x169;
-  out1[59] = x171;
-  out1[60] = x173;
-  out1[61] = x175;
-  out1[62] = x177;
-  out1[63] = x179;
-  out1[64] = x181;
-  out1[65] = x180;
+  out1[28] = x103;
+  out1[29] = x106;
+  out1[30] = x108;
+  out1[31] = x110;
+  out1[32] = x112;
+  out1[33] = x114;
+  out1[34] = x116;
+  out1[35] = x118;
+  out1[36] = x121;
+  out1[37] = x123;
+  out1[38] = x125;
+  out1[39] = x127;
+  out1[40] = x129;
+  out1[41] = x131;
+  out1[42] = x133;
+  out1[43] = x136;
+  out1[44] = x138;
+  out1[45] = x140;
+  out1[46] = x142;
+  out1[47] = x144;
+  out1[48] = x146;
+  out1[49] = x148;
+  out1[50] = x151;
+  out1[51] = x153;
+  out1[52] = x155;
+  out1[53] = x157;
+  out1[54] = x159;
+  out1[55] = x161;
+  out1[56] = x163;
+  out1[57] = x164;
+  out1[58] = x167;
+  out1[59] = x169;
+  out1[60] = x171;
+  out1[61] = x173;
+  out1[62] = x175;
+  out1[63] = x177;
+  out1[64] = x179;
+  out1[65] = x178;
 }
 
 /*
@@ -884,29 +882,27 @@ static void fiat_p521_from_bytes(uint64_t out1[9], const uint8_t arg1[66]) {
   uint8_t x82 = (uint8_t)(x81 >> 58);
   uint64_t x83 = (x81 & UINT64_C(0x3ffffffffffffff));
   uint64_t x84 = (x82 + x75);
-  fiat_p521_uint1 x85 = (fiat_p521_uint1)(x84 >> 58);
-  uint64_t x86 = (x84 & UINT64_C(0x3ffffffffffffff));
-  uint64_t x87 = (x85 + x74);
-  uint8_t x88 = (uint8_t)(x87 >> 58);
-  uint64_t x89 = (x87 & UINT64_C(0x3ffffffffffffff));
-  uint64_t x90 = (x88 + x73);
-  uint8_t x91 = (uint8_t)(x90 >> 58);
-  uint64_t x92 = (x90 & UINT64_C(0x3ffffffffffffff));
-  uint64_t x93 = (x91 + x72);
-  uint8_t x94 = (uint8_t)(x93 >> 58);
-  uint64_t x95 = (x93 & UINT64_C(0x3ffffffffffffff));
-  uint64_t x96 = (x94 + x71);
-  fiat_p521_uint1 x97 = (fiat_p521_uint1)(x96 >> 58);
-  uint64_t x98 = (x96 & UINT64_C(0x3ffffffffffffff));
-  uint64_t x99 = (x97 + x70);
+  uint64_t x85 = (x84 & UINT64_C(0x3ffffffffffffff));
+  uint64_t x86 = (0x0 + x74);
+  uint8_t x87 = (uint8_t)(x86 >> 58);
+  uint64_t x88 = (x86 & UINT64_C(0x3ffffffffffffff));
+  uint64_t x89 = (x87 + x73);
+  uint8_t x90 = (uint8_t)(x89 >> 58);
+  uint64_t x91 = (x89 & UINT64_C(0x3ffffffffffffff));
+  uint64_t x92 = (x90 + x72);
+  uint8_t x93 = (uint8_t)(x92 >> 58);
+  uint64_t x94 = (x92 & UINT64_C(0x3ffffffffffffff));
+  uint64_t x95 = (x93 + x71);
+  uint64_t x96 = (x95 & UINT64_C(0x3ffffffffffffff));
+  uint64_t x97 = (0x0 + x70);
   out1[0] = x69;
   out1[1] = x80;
   out1[2] = x83;
-  out1[3] = x86;
-  out1[4] = x89;
-  out1[5] = x92;
-  out1[6] = x95;
-  out1[7] = x98;
-  out1[8] = x99;
+  out1[3] = x85;
+  out1[4] = x88;
+  out1[5] = x91;
+  out1[6] = x94;
+  out1[7] = x96;
+  out1[8] = x97;
 }