LibCrypto: Move large functions to cpp file

If they use up so much stack space, contain (sometimes several) loops, and take
a noticable amount of time anyway, then 'inline' is probably going to be ignored
by the compiler anyway.

3 files changed, 371 insertions, 326 deletions

diff --git a/Libraries/LibCrypto/CMakeLists.txt b/Libraries/LibCrypto/CMakeLists.txt
index 428f99dc1c..01ddd873ae 100644
--- a/Libraries/LibCrypto/CMakeLists.txt
+++ b/Libraries/LibCrypto/CMakeLists.txt
@@ -1,12 +1,13 @@
 set(SOURCES
-    BigInt/UnsignedBigInteger.cpp
     BigInt/SignedBigInteger.cpp
+    BigInt/UnsignedBigInteger.cpp
     Checksum/Adler32.cpp
     Checksum/CRC32.cpp
     Cipher/AES.cpp
     Hash/MD5.cpp
     Hash/SHA1.cpp
     Hash/SHA2.cpp
+    NumberTheory/ModularFunctions.cpp
     PK/RSA.cpp
 )
 
diff --git a/Libraries/LibCrypto/NumberTheory/ModularFunctions.cpp b/Libraries/LibCrypto/NumberTheory/ModularFunctions.cpp
new file mode 100644
index 0000000000..7386277d7f
--- /dev/null
+++ b/Libraries/LibCrypto/NumberTheory/ModularFunctions.cpp
@@ -0,0 +1,362 @@
+/*
+ * Copyright (c) 2020, Ali Mohammad Pur <ali.mpfard@gmail.com>
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice, this
+ *    list of conditions and the following disclaimer.
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ *    this list of conditions and the following disclaimer in the documentation
+ *    and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+ * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+ * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <LibCrypto/NumberTheory/ModularFunctions.h>
+
+namespace Crypto {
+namespace NumberTheory {
+
+UnsignedBigInteger ModularInverse(const UnsignedBigInteger& a_, const UnsignedBigInteger& b)
+{
+    if (b == 1)
+        return { 1 };
+
+    UnsignedBigInteger one { 1 };
+    UnsignedBigInteger temp_1;
+    UnsignedBigInteger temp_2;
+    UnsignedBigInteger temp_3;
+    UnsignedBigInteger temp_4;
+    UnsignedBigInteger temp_plus;
+    UnsignedBigInteger temp_minus;
+    UnsignedBigInteger temp_quotient;
+    UnsignedBigInteger temp_remainder;
+    UnsignedBigInteger d;
+
+    auto a = a_;
+    auto u = a;
+    if (a.words()[0] % 2 == 0) {
+        // u += b
+        UnsignedBigInteger::add_without_allocation(u, b, temp_plus);
+        u.set_to(temp_plus);
+    }
+
+    auto v = b;
+    UnsignedBigInteger x { 0 };
+
+    // d = b - 1
+    UnsignedBigInteger::subtract_without_allocation(b, one, d);
+
+    while (!(v == 1)) {
+        while (v < u) {
+            // u -= v
+            UnsignedBigInteger::subtract_without_allocation(u, v, temp_minus);
+            u.set_to(temp_minus);
+
+            // d += x
+            UnsignedBigInteger::add_without_allocation(d, x, temp_plus);
+            d.set_to(temp_plus);
+
+            while (u.words()[0] % 2 == 0) {
+                if (d.words()[0] % 2 == 1) {
+                    // d += b
+                    UnsignedBigInteger::add_without_allocation(d, b, temp_plus);
+                    d.set_to(temp_plus);
+                }
+
+                // u /= 2
+                UnsignedBigInteger::divide_u16_without_allocation(u, 2, temp_quotient, temp_remainder);
+                u.set_to(temp_quotient);
+
+                // d /= 2
+                UnsignedBigInteger::divide_u16_without_allocation(d, 2, temp_quotient, temp_remainder);
+                d.set_to(temp_quotient);
+            }
+        }
+
+        // v -= u
+        UnsignedBigInteger::subtract_without_allocation(v, u, temp_minus);
+        v.set_to(temp_minus);
+
+        // x += d
+        UnsignedBigInteger::add_without_allocation(x, d, temp_plus);
+        x.set_to(temp_plus);
+
+        while (v.words()[0] % 2 == 0) {
+            if (x.words()[0] % 2 == 1) {
+                // x += b
+                UnsignedBigInteger::add_without_allocation(x, b, temp_plus);
+                x.set_to(temp_plus);
+            }
+
+            // v /= 2
+            UnsignedBigInteger::divide_u16_without_allocation(v, 2, temp_quotient, temp_remainder);
+            v.set_to(temp_quotient);
+
+            // x /= 2
+            UnsignedBigInteger::divide_u16_without_allocation(x, 2, temp_quotient, temp_remainder);
+            x.set_to(temp_quotient);
+        }
+    }
+
+    // x % b
+    UnsignedBigInteger::divide_without_allocation(x, b, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder);
+    return temp_remainder;
+}
+
+UnsignedBigInteger ModularPower(const UnsignedBigInteger& b, const UnsignedBigInteger& e, const UnsignedBigInteger& m)
+{
+    if (m == 1)
+        return 0;
+
+    UnsignedBigInteger ep { e };
+    UnsignedBigInteger base { b };
+    UnsignedBigInteger exp { 1 };
+
+    UnsignedBigInteger temp_1;
+    UnsignedBigInteger temp_2;
+    UnsignedBigInteger temp_3;
+    UnsignedBigInteger temp_4;
+    UnsignedBigInteger temp_multiply;
+    UnsignedBigInteger temp_quotient;
+    UnsignedBigInteger temp_remainder;
+
+    while (!(ep < 1)) {
+#ifdef NT_DEBUG
+        dbg() << ep.to_base10();
+#endif
+        if (ep.words()[0] % 2 == 1) {
+            // exp = (exp * base) % m;
+            UnsignedBigInteger::multiply_without_allocation(exp, base, temp_1, temp_2, temp_3, temp_4, temp_multiply);
+            UnsignedBigInteger::divide_without_allocation(temp_multiply, m, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder);
+            exp.set_to(temp_remainder);
+        }
+
+        // ep = ep / 2;
+        UnsignedBigInteger::divide_u16_without_allocation(ep, 2, temp_quotient, temp_remainder);
+        ep.set_to(temp_quotient);
+
+        // base = (base * base) % m;
+        UnsignedBigInteger::multiply_without_allocation(base, base, temp_1, temp_2, temp_3, temp_4, temp_multiply);
+        UnsignedBigInteger::divide_without_allocation(temp_multiply, m, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder);
+        base.set_to(temp_remainder);
+    }
+    return exp;
+}
+
+static void GCD_without_allocation(
+    const UnsignedBigInteger& a,
+    const UnsignedBigInteger& b,
+    UnsignedBigInteger& temp_a,
+    UnsignedBigInteger& temp_b,
+    UnsignedBigInteger& temp_1,
+    UnsignedBigInteger& temp_2,
+    UnsignedBigInteger& temp_3,
+    UnsignedBigInteger& temp_4,
+    UnsignedBigInteger& temp_quotient,
+    UnsignedBigInteger& temp_remainder,
+    UnsignedBigInteger& output)
+{
+    temp_a.set_to(a);
+    temp_b.set_to(b);
+    for (;;) {
+        if (temp_a == 0) {
+            output.set_to(temp_b);
+            return;
+        }
+
+        // temp_b %= temp_a
+        UnsignedBigInteger::divide_without_allocation(temp_b, temp_a, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder);
+        temp_b.set_to(temp_remainder);
+        if (temp_b == 0) {
+            output.set_to(temp_a);
+            return;
+        }
+
+        // temp_a %= temp_b
+        UnsignedBigInteger::divide_without_allocation(temp_a, temp_b, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder);
+        temp_a.set_to(temp_remainder);
+    }
+}
+
+UnsignedBigInteger GCD(const UnsignedBigInteger& a, const UnsignedBigInteger& b)
+{
+    UnsignedBigInteger temp_a;
+    UnsignedBigInteger temp_b;
+    UnsignedBigInteger temp_1;
+    UnsignedBigInteger temp_2;
+    UnsignedBigInteger temp_3;
+    UnsignedBigInteger temp_4;
+    UnsignedBigInteger temp_quotient;
+    UnsignedBigInteger temp_remainder;
+    UnsignedBigInteger output;
+
+    GCD_without_allocation(a, b, temp_a, temp_b, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder, output);
+
+    return output;
+}
+
+UnsignedBigInteger LCM(const UnsignedBigInteger& a, const UnsignedBigInteger& b)
+{
+    UnsignedBigInteger temp_a;
+    UnsignedBigInteger temp_b;
+    UnsignedBigInteger temp_1;
+    UnsignedBigInteger temp_2;
+    UnsignedBigInteger temp_3;
+    UnsignedBigInteger temp_4;
+    UnsignedBigInteger temp_quotient;
+    UnsignedBigInteger temp_remainder;
+    UnsignedBigInteger gcd_output;
+    UnsignedBigInteger output { 0 };
+
+    GCD_without_allocation(a, b, temp_a, temp_b, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder, gcd_output);
+    if (gcd_output == 0) {
+#ifdef NT_DEBUG
+        dbg() << "GCD is zero";
+#endif
+        return output;
+    }
+
+    // output = (a / gcd_output) * b
+    UnsignedBigInteger::divide_without_allocation(a, gcd_output, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder);
+    UnsignedBigInteger::multiply_without_allocation(temp_quotient, b, temp_1, temp_2, temp_3, temp_4, output);
+
+#ifdef NT_DEBUG
+    dbg() << "quot: " << temp_quotient << " rem: " << temp_remainder << " out: " << output;
+#endif
+
+    return output;
+}
+
+static bool MR_primality_test(UnsignedBigInteger n, const Vector<UnsignedBigInteger, 256>& tests)
+{
+    // Written using Wikipedia:
+    // https://en.wikipedia.org/wiki/Miller%E2%80%93Rabin_primality_test#Miller%E2%80%93Rabin_test
+    ASSERT(!(n < 4));
+    auto predecessor = n.minus({ 1 });
+    auto d = predecessor;
+    size_t r = 0;
+
+    {
+        auto div_result = d.divided_by(2);
+        while (div_result.remainder == 0) {
+            d = div_result.quotient;
+            div_result = d.divided_by(2);
+            ++r;
+        }
+    }
+    if (r == 0) {
+        // n - 1 is odd, so n was even. But there is only one even prime:
+        return n == 2;
+    }
+
+    for (auto a : tests) {
+        // Technically: ASSERT(2 <= a && a <= n - 2)
+        ASSERT(a < n);
+        auto x = ModularPower(a, d, n);
+        if (x == 1 || x == predecessor)
+            continue;
+        bool skip_this_witness = false;
+        // r − 1 iterations.
+        for (size_t i = 0; i < r - 1; ++i) {
+            x = ModularPower(x, 2, n);
+            if (x == predecessor) {
+                skip_this_witness = true;
+                break;
+            }
+        }
+        if (skip_this_witness)
+            continue;
+        return false; // "composite"
+    }
+
+    return true; // "probably prime"
+}
+
+UnsignedBigInteger random_number(const UnsignedBigInteger& min, const UnsignedBigInteger& max_excluded)
+{
+    ASSERT(min < max_excluded);
+    auto range = max_excluded.minus(min);
+    UnsignedBigInteger base;
+    auto size = range.trimmed_length() * sizeof(u32) + 2;
+    // "+2" is intentional (see below).
+    // Also, if we're about to crash anyway, at least produce a nice error:
+    ASSERT(size < 8 * MB);
+    u8 buf[size];
+    AK::fill_with_random(buf, size);
+    UnsignedBigInteger random { buf, size };
+    // At this point, `random` is a large number, in the range [0, 256^size).
+    // To get down to the actual range, we could just compute random % range.
+    // This introduces "modulo bias". However, since we added 2 to `size`,
+    // we know that the generated range is at least 65536 times as large as the
+    // required range! This means that the modulo bias is only 0.0015%, if all
+    // inputs are chosen adversarially. Let's hope this is good enough.
+    auto divmod = random.divided_by(range);
+    // The proper way to fix this is to restart if `divmod.quotient` is maximal.
+    return divmod.remainder.plus(min);
+}
+
+bool is_probably_prime(const UnsignedBigInteger& p)
+{
+    // Is it a small number?
+    if (p < 49) {
+        u32 p_value = p.words()[0];
+        // Is it a very small prime?
+        if (p_value == 2 || p_value == 3 || p_value == 5 || p_value == 7)
+            return true;
+        // Is it the multiple of a very small prime?
+        if (p_value % 2 == 0 || p_value % 3 == 0 || p_value % 5 == 0 || p_value % 7 == 0)
+            return false;
+        // Then it must be a prime, but not a very small prime, like 37.
+        return true;
+    }
+
+    Vector<UnsignedBigInteger, 256> tests;
+    // Make some good initial guesses that are guaranteed to find all primes < 2^64.
+    tests.append(UnsignedBigInteger(2));
+    tests.append(UnsignedBigInteger(3));
+    tests.append(UnsignedBigInteger(5));
+    tests.append(UnsignedBigInteger(7));
+    tests.append(UnsignedBigInteger(11));
+    tests.append(UnsignedBigInteger(13));
+    UnsignedBigInteger seventeen { 17 };
+    for (size_t i = tests.size(); i < 256; ++i) {
+        tests.append(random_number(seventeen, p.minus(2)));
+    }
+    // Miller-Rabin's "error" is 8^-k. In adversarial cases, it's 4^-k.
+    // With 200 random numbers, this would mean an error of about 2^-400.
+    // So we don't need to worry too much about the quality of the random numbers.
+
+    return MR_primality_test(p, tests);
+}
+
+UnsignedBigInteger random_big_prime(size_t bits)
+{
+    ASSERT(bits >= 33);
+    UnsignedBigInteger min = UnsignedBigInteger::from_base10("6074001000").shift_left(bits - 33);
+    UnsignedBigInteger max = UnsignedBigInteger { 1 }.shift_left(bits).minus(1);
+    for (;;) {
+        auto p = random_number(min, max);
+        if ((p.words()[0] & 1) == 0) {
+            // An even number is definitely not a large prime.
+            continue;
+        }
+        if (is_probably_prime(p))
+            return p;
+    }
+}
+
+}
+}
diff --git a/Libraries/LibCrypto/NumberTheory/ModularFunctions.h b/Libraries/LibCrypto/NumberTheory/ModularFunctions.h
index 53699701af..586a5b3e10 100644
--- a/Libraries/LibCrypto/NumberTheory/ModularFunctions.h
+++ b/Libraries/LibCrypto/NumberTheory/ModularFunctions.h
@@ -34,132 +34,8 @@
 namespace Crypto {
 namespace NumberTheory {
 
-inline UnsignedBigInteger ModularInverse(const UnsignedBigInteger& a_, const UnsignedBigInteger& b)
-{
-    if (b == 1)
-        return { 1 };
-
-    UnsignedBigInteger one { 1 };
-    UnsignedBigInteger temp_1;
-    UnsignedBigInteger temp_2;
-    UnsignedBigInteger temp_3;
-    UnsignedBigInteger temp_4;
-    UnsignedBigInteger temp_plus;
-    UnsignedBigInteger temp_minus;
-    UnsignedBigInteger temp_quotient;
-    UnsignedBigInteger temp_remainder;
-    UnsignedBigInteger d;
-
-    auto a = a_;
-    auto u = a;
-    if (a.words()[0] % 2 == 0) {
-        // u += b
-        UnsignedBigInteger::add_without_allocation(u, b, temp_plus);
-        u.set_to(temp_plus);
-    }
-
-    auto v = b;
-    UnsignedBigInteger x { 0 };
-
-    // d = b - 1
-    UnsignedBigInteger::subtract_without_allocation(b, one, d);
-
-    while (!(v == 1)) {
-        while (v < u) {
-            // u -= v
-            UnsignedBigInteger::subtract_without_allocation(u, v, temp_minus);
-            u.set_to(temp_minus);
-
-            // d += x
-            UnsignedBigInteger::add_without_allocation(d, x, temp_plus);
-            d.set_to(temp_plus);
-
-            while (u.words()[0] % 2 == 0) {
-                if (d.words()[0] % 2 == 1) {
-                    // d += b
-                    UnsignedBigInteger::add_without_allocation(d, b, temp_plus);
-                    d.set_to(temp_plus);
-                }
-
-                // u /= 2
-                UnsignedBigInteger::divide_u16_without_allocation(u, 2, temp_quotient, temp_remainder);
-                u.set_to(temp_quotient);
-
-                // d /= 2
-                UnsignedBigInteger::divide_u16_without_allocation(d, 2, temp_quotient, temp_remainder);
-                d.set_to(temp_quotient);
-            }
-        }
-
-        // v -= u
-        UnsignedBigInteger::subtract_without_allocation(v, u, temp_minus);
-        v.set_to(temp_minus);
-
-        // x += d
-        UnsignedBigInteger::add_without_allocation(x, d, temp_plus);
-        x.set_to(temp_plus);
-
-        while (v.words()[0] % 2 == 0) {
-            if (x.words()[0] % 2 == 1) {
-                // x += b
-                UnsignedBigInteger::add_without_allocation(x, b, temp_plus);
-                x.set_to(temp_plus);
-            }
-
-            // v /= 2
-            UnsignedBigInteger::divide_u16_without_allocation(v, 2, temp_quotient, temp_remainder);
-            v.set_to(temp_quotient);
-
-            // x /= 2
-            UnsignedBigInteger::divide_u16_without_allocation(x, 2, temp_quotient, temp_remainder);
-            x.set_to(temp_quotient);
-        }
-    }
-
-    // x % b
-    UnsignedBigInteger::divide_without_allocation(x, b, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder);
-    return temp_remainder;
-}
-
-static UnsignedBigInteger ModularPower(const UnsignedBigInteger& b, const UnsignedBigInteger& e, const UnsignedBigInteger& m)
-{
-    if (m == 1)
-        return 0;
-
-    UnsignedBigInteger ep { e };
-    UnsignedBigInteger base { b };
-    UnsignedBigInteger exp { 1 };
-
-    UnsignedBigInteger temp_1;
-    UnsignedBigInteger temp_2;
-    UnsignedBigInteger temp_3;
-    UnsignedBigInteger temp_4;
-    UnsignedBigInteger temp_multiply;
-    UnsignedBigInteger temp_quotient;
-    UnsignedBigInteger temp_remainder;
-
-    while (!(ep < 1)) {
-#ifdef NT_DEBUG
-        dbg() << ep.to_base10();
-#endif
-        if (ep.words()[0] % 2 == 1) {
-            // exp = (exp * base) % m;
-            UnsignedBigInteger::multiply_without_allocation(exp, base, temp_1, temp_2, temp_3, temp_4, temp_multiply);
-            UnsignedBigInteger::divide_without_allocation(temp_multiply, m, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder);
-            exp.set_to(temp_remainder);
-        }
-
-        // ep = ep / 2;
-        UnsignedBigInteger::divide_u16_without_allocation(ep, 2, temp_quotient, temp_remainder);
-        ep.set_to(temp_quotient);
-
-        // base = (base * base) % m;
-        UnsignedBigInteger::multiply_without_allocation(base, base, temp_1, temp_2, temp_3, temp_4, temp_multiply);
-        UnsignedBigInteger::divide_without_allocation(temp_multiply, m, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder);
-        base.set_to(temp_remainder);
-    }
-    return exp;
-}
+UnsignedBigInteger ModularInverse(const UnsignedBigInteger& a_, const UnsignedBigInteger& b);
+UnsignedBigInteger ModularPower(const UnsignedBigInteger& b, const UnsignedBigInteger& e, const UnsignedBigInteger& m);
 
 // Note: This function _will_ generate extremely huge numbers, and in doing so,
 //       it will allocate and free a lot of memory!
@@ -185,206 +61,12 @@ static IntegerType Power(const IntegerType& b, const IntegerType& e)
     return exp;
 }
 
-static void GCD_without_allocation(
-    const UnsignedBigInteger& a,
-    const UnsignedBigInteger& b,
-    UnsignedBigInteger& temp_a,
-    UnsignedBigInteger& temp_b,
-    UnsignedBigInteger& temp_1,
-    UnsignedBigInteger& temp_2,
-    UnsignedBigInteger& temp_3,
-    UnsignedBigInteger& temp_4,
-    UnsignedBigInteger& temp_quotient,
-    UnsignedBigInteger& temp_remainder,
-    UnsignedBigInteger& output)
-{
-    temp_a.set_to(a);
-    temp_b.set_to(b);
-    for (;;) {
-        if (temp_a == 0) {
-            output.set_to(temp_b);
-            return;
-        }
-
-        // temp_b %= temp_a
-        UnsignedBigInteger::divide_without_allocation(temp_b, temp_a, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder);
-        temp_b.set_to(temp_remainder);
-        if (temp_b == 0) {
-            output.set_to(temp_a);
-            return;
-        }
-
-        // temp_a %= temp_b
-        UnsignedBigInteger::divide_without_allocation(temp_a, temp_b, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder);
-        temp_a.set_to(temp_remainder);
-    }
-}
-
-inline UnsignedBigInteger GCD(const UnsignedBigInteger& a, const UnsignedBigInteger& b)
-{
-    UnsignedBigInteger temp_a;
-    UnsignedBigInteger temp_b;
-    UnsignedBigInteger temp_1;
-    UnsignedBigInteger temp_2;
-    UnsignedBigInteger temp_3;
-    UnsignedBigInteger temp_4;
-    UnsignedBigInteger temp_quotient;
-    UnsignedBigInteger temp_remainder;
-    UnsignedBigInteger output;
-
-    GCD_without_allocation(a, b, temp_a, temp_b, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder, output);
-
-    return output;
-}
-
-inline UnsignedBigInteger LCM(const UnsignedBigInteger& a, const UnsignedBigInteger& b)
-{
-    UnsignedBigInteger temp_a;
-    UnsignedBigInteger temp_b;
-    UnsignedBigInteger temp_1;
-    UnsignedBigInteger temp_2;
-    UnsignedBigInteger temp_3;
-    UnsignedBigInteger temp_4;
-    UnsignedBigInteger temp_quotient;
-    UnsignedBigInteger temp_remainder;
-    UnsignedBigInteger gcd_output;
-    UnsignedBigInteger output { 0 };
-
-    GCD_without_allocation(a, b, temp_a, temp_b, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder, gcd_output);
-    if (gcd_output == 0) {
-#ifdef NT_DEBUG
-        dbg() << "GCD is zero";
-#endif
-        return output;
-    }
-
-    // output = (a / gcd_output) * b
-    UnsignedBigInteger::divide_without_allocation(a, gcd_output, temp_1, temp_2, temp_3, temp_4, temp_quotient, temp_remainder);
-    UnsignedBigInteger::multiply_without_allocation(temp_quotient, b, temp_1, temp_2, temp_3, temp_4, output);
-
-#ifdef NT_DEBUG
-    dbg() << "quot: " << temp_quotient << " rem: " << temp_remainder << " out: " << output;
-#endif
-
-    return output;
-}
-
-template<size_t test_count>
-static bool MR_primality_test(UnsignedBigInteger n, const Vector<UnsignedBigInteger, test_count>& tests)
-{
-    // Written using Wikipedia:
-    // https://en.wikipedia.org/wiki/Miller%E2%80%93Rabin_primality_test#Miller%E2%80%93Rabin_test
-    ASSERT(!(n < 4));
-    auto predecessor = n.minus({ 1 });
-    auto d = predecessor;
-    size_t r = 0;
-
-    {
-        auto div_result = d.divided_by(2);
-        while (div_result.remainder == 0) {
-            d = div_result.quotient;
-            div_result = d.divided_by(2);
-            ++r;
-        }
-    }
-    if (r == 0) {
-        // n - 1 is odd, so n was even. But there is only one even prime:
-        return n == 2;
-    }
+UnsignedBigInteger GCD(const UnsignedBigInteger& a, const UnsignedBigInteger& b);
+UnsignedBigInteger LCM(const UnsignedBigInteger& a, const UnsignedBigInteger& b);
 
-    for (auto a : tests) {
-        // Technically: ASSERT(2 <= a && a <= n - 2)
-        ASSERT(a < n);
-        auto x = ModularPower(a, d, n);
-        if (x == 1 || x == predecessor)
-            continue;
-        bool skip_this_witness = false;
-        // r − 1 iterations.
-        for (size_t i = 0; i < r - 1; ++i) {
-            x = ModularPower(x, 2, n);
-            if (x == predecessor) {
-                skip_this_witness = true;
-                break;
-            }
-        }
-        if (skip_this_witness)
-            continue;
-        return false; // "composite"
-    }
-
-    return true; // "probably prime"
-}
-
-static UnsignedBigInteger random_number(const UnsignedBigInteger& min, const UnsignedBigInteger& max_excluded)
-{
-    ASSERT(min < max_excluded);
-    auto range = max_excluded.minus(min);
-    UnsignedBigInteger base;
-    auto size = range.trimmed_length() * sizeof(u32) + 2;
-    // "+2" is intentional (see below).
-    u8 buf[size];
-    AK::fill_with_random(buf, size);
-    UnsignedBigInteger random { buf, size };
-    // At this point, `random` is a large number, in the range [0, 256^size).
-    // To get down to the actual range, we could just compute random % range.
-    // This introduces "modulo bias". However, since we added 2 to `size`,
-    // we know that the generated range is at least 65536 times as large as the
-    // required range! This means that the modulo bias is only 0.0015%, if all
-    // inputs are chosen adversarially. Let's hope this is good enough.
-    auto divmod = random.divided_by(range);
-    // The proper way to fix this is to restart if `divmod.quotient` is maximal.
-    return divmod.remainder.plus(min);
-}
-
-static bool is_probably_prime(const UnsignedBigInteger& p)
-{
-    // Is it a small number?
-    if (p < 49) {
-        u32 p_value = p.words()[0];
-        // Is it a very small prime?
-        if (p_value == 2 || p_value == 3 || p_value == 5 || p_value == 7)
-            return true;
-        // Is it the multiple of a very small prime?
-        if (p_value % 2 == 0 || p_value % 3 == 0 || p_value % 5 == 0 || p_value % 7 == 0)
-            return false;
-        // Then it must be a prime, but not a very small prime, like 37.
-        return true;
-    }
-
-    Vector<UnsignedBigInteger, 256> tests;
-    // Make some good initial guesses that are guaranteed to find all primes < 2^64.
-    tests.append(UnsignedBigInteger(2));
-    tests.append(UnsignedBigInteger(3));
-    tests.append(UnsignedBigInteger(5));
-    tests.append(UnsignedBigInteger(7));
-    tests.append(UnsignedBigInteger(11));
-    tests.append(UnsignedBigInteger(13));
-    UnsignedBigInteger seventeen { 17 };
-    for (size_t i = tests.size(); i < 256; ++i) {
-        tests.append(random_number(seventeen, p.minus(2)));
-    }
-    // Miller-Rabin's "error" is 8^-k. In adversarial cases, it's 4^-k.
-    // With 200 random numbers, this would mean an error of about 2^-400.
-    // So we don't need to worry too much about the quality of the random numbers.
-
-    return MR_primality_test(p, tests);
-}
-
-inline static UnsignedBigInteger random_big_prime(size_t bits)
-{
-    ASSERT(bits >= 33);
-    UnsignedBigInteger min = UnsignedBigInteger::from_base10("6074001000").shift_left(bits - 33);
-    UnsignedBigInteger max = UnsignedBigInteger { 1 }.shift_left(bits).minus(1);
-    for (;;) {
-        auto p = random_number(min, max);
-        if ((p.words()[0] & 1) == 0) {
-            // An even number is definitely not a large prime.
-            continue;
-        }
-        if (is_probably_prime(p))
-            return p;
-    }
-}
+UnsignedBigInteger random_number(const UnsignedBigInteger& min, const UnsignedBigInteger& max_excluded);
+bool is_probably_prime(const UnsignedBigInteger& p);
+UnsignedBigInteger random_big_prime(size_t bits);
 
 }
 }