Do the 16 step twice since 32 is not valid

Author: mborland

include/boost/decimal/detail/remove_trailing_zeros.hpp

@@ -95,64 +95,37 @@ constexpr auto remove_trailing_zeros(std::uint64_t n) noexcept -> remove_trailin
     return {n, s};
 }
 
-/*
-constexpr auto remove_trailing_zeros(uint128 n) noexcept -> remove_trailing_zeros_return<uint128>
-{
-    if (n.high == UINT64_C(0))
-    {
-        const auto temp {remove_trailing_zeros(n.low)};
-        return {static_cast<uint128>(temp.trimmed_number), temp.number_of_removed_zeros};
-    }
-
-    std::size_t s = 0;
-    constexpr detail::uint128 m {UINT64_C(0xCCCCCCCCCCCCCCCC), UINT64_C(0xCCCCCCCCCCCCCCCD)};
-    constexpr detail::uint128 threshold_value {UINT64_C(0x1999999999999999), UINT64_C(0x999999999999999A)};
-
-    while (true)
-    {
-        const auto r = rotr<128>(n * m, 1);
-
-        if (r < threshold_value)
-        {
-            n = r;
-            s += 1;
-        }
-        else
-        {
-            break;
-        }
-    }
-
-    return {n, s};
-}
-*/
-
 constexpr auto remove_trailing_zeros(uint128 n) noexcept -> remove_trailing_zeros_return<uint128>
 {
     std::size_t s {};
 
-    auto r = rotr<128>(n * uint128{UINT64_C(0x3275305C1066), UINT64_C(0xE4A4D1417CD9A041)}, 16);
-    auto b = r < uint128{UINT64_C(0x734), UINT64_C(0xACA5F6226F0ADA62)};
+    auto r = rotr<128>(n * uint128 {UINT64_C(0x3275305C1066), UINT64_C(0xE4A4D1417CD9A041)}, 16);
+    auto b = r < uint128 {UINT64_C(0x734), UINT64_C(0xACA5F6226F0ADA62)};
+    s = s * 2U + static_cast<std::size_t>(b);
+    n = b ? r : n;
+
+    r = rotr<128>(n * uint128 {UINT64_C(0x3275305C1066), UINT64_C(0xE4A4D1417CD9A041)}, 16);
+    b = r < uint128 {UINT64_C(0x734), UINT64_C(0xACA5F6226F0ADA62)};
     s = s * 2U + static_cast<std::size_t>(b);
     n = b ? r : n;
 
-    r = rotr<128>(n * uint128{UINT64_C(0x6B7213EE9F5A78), UINT64_C(0xC767074B22E90E21)}, 8);
-    b = r < uint128{UINT64_C(0x2AF31DC461), UINT64_C(0x1873BF3F70834ACE)};
+    r = rotr<128>(n * uint128 {UINT64_C(0x6B7213EE9F5A78), UINT64_C(0xC767074B22E90E21)}, 8);
+    b = r < uint128 {UINT64_C(0x2AF31DC461), UINT64_C(0x1873BF3F70834ACE)};
     s = s * 2U + static_cast<std::size_t>(b);
     n = b ? r : n;
 
-    r = rotr<128>(n * uint128{UINT64_C(0x95182A9930BE0DE), UINT64_C(0xD288CE703AFB7E91)}, 4);
-    b = r < uint128{UINT64_C(0x68DB8BAC710CB), UINT64_C(0x295E9E1B089A0276)};
+    r = rotr<128>(n * uint128 {UINT64_C(0x95182A9930BE0DE), UINT64_C(0xD288CE703AFB7E91)}, 4);
+    b = r < uint128 {UINT64_C(0x68DB8BAC710CB), UINT64_C(0x295E9E1B089A0276)};
     s = s * 2U + static_cast<std::size_t>(b);
     n = b ? r : n;
 
-    r = rotr<128>(n * uint128{UINT64_C(0x28F5C28F5C28F5C2), UINT64_C(0x8F5C28F5C28F5C29)}, 2);
-    b = r < uint128{UINT64_C(0x28F5C28F5C28F5C), UINT64_C(0x28F5C28F5C28F5C3)};
+    r = rotr<128>(n * uint128 {UINT64_C(0x28F5C28F5C28F5C2), UINT64_C(0x8F5C28F5C28F5C29)}, 2);
+    b = r < uint128 {UINT64_C(0x28F5C28F5C28F5C), UINT64_C(0x28F5C28F5C28F5C3)};
     s = s * 2U + static_cast<std::size_t>(b);
     n = b ? r : n;
 
-    r = rotr<128>(n * uint128{UINT64_C(0xCCCCCCCCCCCCCCCC), UINT64_C(0xCCCCCCCCCCCCCCCD)}, 1);
-    b = r < uint128{UINT64_C(0x1999999999999999), UINT64_C(0x999999999999999A)};
+    r = rotr<128>(n * uint128 {UINT64_C(0xCCCCCCCCCCCCCCCC), UINT64_C(0xCCCCCCCCCCCCCCCD)}, 1);
+    b = r < uint128 {UINT64_C(0x1999999999999999), UINT64_C(0x999999999999999A)};
     s = s * 2U + static_cast<std::size_t>(b);
     n = b ? r : n;
 

tools/granlund-montgomery.py

@@ -9,13 +9,13 @@ def extended_euclidean(a, b):
 def mod_inverse(a, m):
     gcd, x, _ = extended_euclidean(a, m)
     if gcd != 1:
-        return None  # Modular inverse doesn't exist
+        return 0  # Modular inverse doesn't exist
     else:
         return x % m
 
 # Constants
 bits = int(128)
-t = int(2)
+t = int(32)
 
 q = int(10)**t
 q0 = int(q / int(2)**t)