Merge #15512: Add ChaCha20 encryption option (XOR)

2dfe275 Add ChaCha20 bench (Jonas Schnelli)
2bc2b8b Add ChaCha20 encryption option (XOR) (Jonas Schnelli)

Pull request description:

  The current ChaCha20 implementation does not support message encryption (it can only output the keystream which is sufficient for the RNG).

  This PR adds the actual XORing of the `plaintext` with the `keystream` in order to return the desired `ciphertext`.

  Required for v2 message transport protocol.

ACKs for commit 2dfe27:
  jnewbery:
    Looks good. utACK 2dfe275.
  jnewbery:
    utACK 2dfe275
  sipa:
    utACK 2dfe275
  ryanofsky:
    utACK 2dfe275. Changes since last review are just renaming the Crypt method, adding comments, and simplifying the benchmark.

Tree-SHA512: 84bb234da2ca9fdc44bc29a786d9dd215520f81245270c1aef801ef66b6091b7793e2eb38ad6dbb084925245065c5dce9e5582f2d0fa220ab3e182d43412d5b5

Author: jonasschnelli

src/Makefile.bench.include

@@ -21,6 +21,7 @@ bench_bench_bitcoin_SOURCES = \
   bench/duplicate_inputs.cpp \
   bench/examples.cpp \
   bench/rollingbloom.cpp \
+  bench/chacha20.cpp \
   bench/crypto_hash.cpp \
   bench/ccoins_caching.cpp \
   bench/gcs_filter.cpp \

src/bench/chacha20.cpp

@@ -0,0 +1,46 @@
+// Copyright (c) 2019 The Bitcoin Core developers
+// Distributed under the MIT software license, see the accompanying
+// file COPYING or http://www.opensource.org/licenses/mit-license.php.
+
+#include <iostream>
+
+#include <bench/bench.h>
+#include <hash.h>
+#include <crypto/chacha20.h>
+
+/* Number of bytes to process per iteration */
+static const uint64_t BUFFER_SIZE_TINY  = 64;
+static const uint64_t BUFFER_SIZE_SMALL = 256;
+static const uint64_t BUFFER_SIZE_LARGE = 1024*1024;
+
+static void CHACHA20(benchmark::State& state, size_t buffersize)
+{
+    std::vector<uint8_t> key(32,0);
+    ChaCha20 ctx(key.data(), key.size());
+    ctx.SetIV(0);
+    ctx.Seek(0);
+    std::vector<uint8_t> in(buffersize,0);
+    std::vector<uint8_t> out(buffersize,0);
+    while (state.KeepRunning()) {
+        ctx.Crypt(in.data(), out.data(), in.size());
+    }
+}
+
+static void CHACHA20_64BYTES(benchmark::State& state)
+{
+    CHACHA20(state, BUFFER_SIZE_TINY);
+}
+
+static void CHACHA20_256BYTES(benchmark::State& state)
+{
+    CHACHA20(state, BUFFER_SIZE_SMALL);
+}
+
+static void CHACHA20_1MB(benchmark::State& state)
+{
+    CHACHA20(state, BUFFER_SIZE_LARGE);
+}
+
+BENCHMARK(CHACHA20_64BYTES, 500000);
+BENCHMARK(CHACHA20_256BYTES, 250000);
+BENCHMARK(CHACHA20_1MB, 340);

src/crypto/chacha20.cpp

@@ -71,7 +71,7 @@ void ChaCha20::Seek(uint64_t pos)
     input[13] = pos >> 32;
 }
 
-void ChaCha20::Output(unsigned char* c, size_t bytes)
+void ChaCha20::Keystream(unsigned char* c, size_t bytes)
 {
     uint32_t x0, x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13, x14, x15;
     uint32_t j0, j1, j2, j3, j4, j5, j6, j7, j8, j9, j10, j11, j12, j13, j14, j15;
@@ -178,3 +178,133 @@ void ChaCha20::Output(unsigned char* c, size_t bytes)
         c += 64;
     }
 }
+
+void ChaCha20::Crypt(const unsigned char* m, unsigned char* c, size_t bytes)
+{
+    uint32_t x0, x1, x2, x3, x4, x5, x6, x7, x8, x9, x10, x11, x12, x13, x14, x15;
+    uint32_t j0, j1, j2, j3, j4, j5, j6, j7, j8, j9, j10, j11, j12, j13, j14, j15;
+    unsigned char *ctarget = nullptr;
+    unsigned char tmp[64];
+    unsigned int i;
+
+    if (!bytes) return;
+
+    j0 = input[0];
+    j1 = input[1];
+    j2 = input[2];
+    j3 = input[3];
+    j4 = input[4];
+    j5 = input[5];
+    j6 = input[6];
+    j7 = input[7];
+    j8 = input[8];
+    j9 = input[9];
+    j10 = input[10];
+    j11 = input[11];
+    j12 = input[12];
+    j13 = input[13];
+    j14 = input[14];
+    j15 = input[15];
+
+    for (;;) {
+        if (bytes < 64) {
+            // if m has fewer than 64 bytes available, copy m to tmp and
+            // read from tmp instead
+            for (i = 0;i < bytes;++i) tmp[i] = m[i];
+            m = tmp;
+            ctarget = c;
+            c = tmp;
+        }
+        x0 = j0;
+        x1 = j1;
+        x2 = j2;
+        x3 = j3;
+        x4 = j4;
+        x5 = j5;
+        x6 = j6;
+        x7 = j7;
+        x8 = j8;
+        x9 = j9;
+        x10 = j10;
+        x11 = j11;
+        x12 = j12;
+        x13 = j13;
+        x14 = j14;
+        x15 = j15;
+        for (i = 20;i > 0;i -= 2) {
+            QUARTERROUND( x0, x4, x8,x12)
+            QUARTERROUND( x1, x5, x9,x13)
+            QUARTERROUND( x2, x6,x10,x14)
+            QUARTERROUND( x3, x7,x11,x15)
+            QUARTERROUND( x0, x5,x10,x15)
+            QUARTERROUND( x1, x6,x11,x12)
+            QUARTERROUND( x2, x7, x8,x13)
+            QUARTERROUND( x3, x4, x9,x14)
+        }
+        x0 += j0;
+        x1 += j1;
+        x2 += j2;
+        x3 += j3;
+        x4 += j4;
+        x5 += j5;
+        x6 += j6;
+        x7 += j7;
+        x8 += j8;
+        x9 += j9;
+        x10 += j10;
+        x11 += j11;
+        x12 += j12;
+        x13 += j13;
+        x14 += j14;
+        x15 += j15;
+
+        x0 ^= ReadLE32(m + 0);
+        x1 ^= ReadLE32(m + 4);
+        x2 ^= ReadLE32(m + 8);
+        x3 ^= ReadLE32(m + 12);
+        x4 ^= ReadLE32(m + 16);
+        x5 ^= ReadLE32(m + 20);
+        x6 ^= ReadLE32(m + 24);
+        x7 ^= ReadLE32(m + 28);
+        x8 ^= ReadLE32(m + 32);
+        x9 ^= ReadLE32(m + 36);
+        x10 ^= ReadLE32(m + 40);
+        x11 ^= ReadLE32(m + 44);
+        x12 ^= ReadLE32(m + 48);
+        x13 ^= ReadLE32(m + 52);
+        x14 ^= ReadLE32(m + 56);
+        x15 ^= ReadLE32(m + 60);
+
+        ++j12;
+        if (!j12) ++j13;
+
+        WriteLE32(c + 0, x0);
+        WriteLE32(c + 4, x1);
+        WriteLE32(c + 8, x2);
+        WriteLE32(c + 12, x3);
+        WriteLE32(c + 16, x4);
+        WriteLE32(c + 20, x5);
+        WriteLE32(c + 24, x6);
+        WriteLE32(c + 28, x7);
+        WriteLE32(c + 32, x8);
+        WriteLE32(c + 36, x9);
+        WriteLE32(c + 40, x10);
+        WriteLE32(c + 44, x11);
+        WriteLE32(c + 48, x12);
+        WriteLE32(c + 52, x13);
+        WriteLE32(c + 56, x14);
+        WriteLE32(c + 60, x15);
+
+        if (bytes <= 64) {
+            if (bytes < 64) {
+                for (i = 0;i < bytes;++i) ctarget[i] = c[i];
+            }
+            input[12] = j12;
+            input[13] = j13;
+            return;
+        }
+        bytes -= 64;
+        c += 64;
+        m += 64;
+    }
+}

src/crypto/chacha20.h

@@ -8,7 +8,8 @@
 #include <stdint.h>
 #include <stdlib.h>
 
-/** A PRNG class for ChaCha20. */
+/** A class for ChaCha20 256-bit stream cipher developed by Daniel J. Bernstein
+    https://cr.yp.to/chacha/chacha-20080128.pdf */
 class ChaCha20
 {
 private:
@@ -17,10 +18,17 @@ class ChaCha20
 public:
     ChaCha20();
     ChaCha20(const unsigned char* key, size_t keylen);
-    void SetKey(const unsigned char* key, size_t keylen);
-    void SetIV(uint64_t iv);
-    void Seek(uint64_t pos);
-    void Output(unsigned char* output, size_t bytes);
+    void SetKey(const unsigned char* key, size_t keylen); //!< set key with flexible keylength; 256bit recommended */
+    void SetIV(uint64_t iv); // set the 64bit nonce
+    void Seek(uint64_t pos); // set the 64bit block counter
+
+    /** outputs the keystream of size <bytes> into <c> */
+    void Keystream(unsigned char* c, size_t bytes);
+
+    /** enciphers the message <input> of length <bytes> and write the enciphered representation into <output>
+     *  Used for encryption and decryption (XOR)
+     */
+    void Crypt(const unsigned char* input, unsigned char* output, size_t bytes);
 };
 
 #endif // BITCOIN_CRYPTO_CHACHA20_H

src/random.cpp

@@ -652,7 +652,7 @@ std::vector<unsigned char> FastRandomContext::randbytes(size_t len)
     if (requires_seed) RandomSeed();
     std::vector<unsigned char> ret(len);
     if (len > 0) {
-        rng.Output(&ret[0], len);
+        rng.Keystream(&ret[0], len);
     }
     return ret;
 }

src/random.h

@@ -111,7 +111,7 @@ class FastRandomContext {
         if (requires_seed) {
             RandomSeed();
         }
-        rng.Output(bytebuf, sizeof(bytebuf));
+        rng.Keystream(bytebuf, sizeof(bytebuf));
         bytebuf_size = sizeof(bytebuf);
     }
 

src/test/crypto_tests.cpp

@@ -125,17 +125,36 @@ static void TestAES256CBC(const std::string &hexkey, const std::string &hexiv, b
     }
 }
 
-static void TestChaCha20(const std::string &hexkey, uint64_t nonce, uint64_t seek, const std::string& hexout)
+static void TestChaCha20(const std::string &hex_message, const std::string &hexkey, uint64_t nonce, uint64_t seek, const std::string& hexout)
 {
     std::vector<unsigned char> key = ParseHex(hexkey);
+    std::vector<unsigned char> m = ParseHex(hex_message);
     ChaCha20 rng(key.data(), key.size());
     rng.SetIV(nonce);
     rng.Seek(seek);
     std::vector<unsigned char> out = ParseHex(hexout);
     std::vector<unsigned char> outres;
     outres.resize(out.size());
-    rng.Output(outres.data(), outres.size());
+    assert(hex_message.empty() || m.size() == out.size());
+
+    // perform the ChaCha20 round(s), if message is provided it will output the encrypted ciphertext otherwise the keystream
+    if (!hex_message.empty()) {
+        rng.Crypt(m.data(), outres.data(), outres.size());
+    } else {
+        rng.Keystream(outres.data(), outres.size());
+    }
     BOOST_CHECK(out == outres);
+    if (!hex_message.empty()) {
+        // Manually XOR with the keystream and compare the output
+        rng.SetIV(nonce);
+        rng.Seek(seek);
+        std::vector<unsigned char> only_keystream(outres.size());
+        rng.Keystream(only_keystream.data(), only_keystream.size());
+        for (size_t i = 0; i != m.size(); i++) {
+            outres[i] = m[i] ^ only_keystream[i];
+        }
+        BOOST_CHECK(out == outres);
+    }
 }
 
 static void TestPoly1305(const std::string &hexmessage, const std::string &hexkey, const std::string& hextag)
@@ -420,25 +439,37 @@ BOOST_AUTO_TEST_CASE(aes_cbc_testvectors) {
 BOOST_AUTO_TEST_CASE(chacha20_testvector)
 {
     // Test vector from RFC 7539
-    TestChaCha20("000102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1f", 0x4a000000UL, 1,
+
+    // test encryption
+    TestChaCha20("4c616469657320616e642047656e746c656d656e206f662074686520636c617373206f66202739393a204966204920636f756"
+                 "c64206f6666657220796f75206f6e6c79206f6e652074697020666f7220746865206675747572652c2073756e73637265656e"
+                 "20776f756c642062652069742e",
+                 "000102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1f", 0x4a000000UL, 1,
+                 "6e2e359a2568f98041ba0728dd0d6981e97e7aec1d4360c20a27afccfd9fae0bf91b65c5524733ab8f593dabcd62b3571639d"
+                 "624e65152ab8f530c359f0861d807ca0dbf500d6a6156a38e088a22b65e52bc514d16ccf806818ce91ab77937365af90bbf74"
+                 "a35be6b40b8eedf2785e42874d"
+                 );
+
+    // test keystream output
+    TestChaCha20("", "000102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1f", 0x4a000000UL, 1,
                  "224f51f3401bd9e12fde276fb8631ded8c131f823d2c06e27e4fcaec9ef3cf788a3b0aa372600a92b57974cded2b9334794cb"
                  "a40c63e34cdea212c4cf07d41b769a6749f3f630f4122cafe28ec4dc47e26d4346d70b98c73f3e9c53ac40c5945398b6eda1a"
                  "832c89c167eacd901d7e2bf363");
 
     // Test vectors from https://tools.ietf.org/html/draft-agl-tls-chacha20poly1305-04#section-7
-    TestChaCha20("0000000000000000000000000000000000000000000000000000000000000000", 0, 0,
+    TestChaCha20("", "0000000000000000000000000000000000000000000000000000000000000000", 0, 0,
                  "76b8e0ada0f13d90405d6ae55386bd28bdd219b8a08ded1aa836efcc8b770dc7da41597c5157488d7724e03fb8d84a376a43b"
                  "8f41518a11cc387b669b2ee6586");
-    TestChaCha20("0000000000000000000000000000000000000000000000000000000000000001", 0, 0,
+    TestChaCha20("", "0000000000000000000000000000000000000000000000000000000000000001", 0, 0,
                  "4540f05a9f1fb296d7736e7b208e3c96eb4fe1834688d2604f450952ed432d41bbe2a0b6ea7566d2a5d1e7e20d42af2c53d79"
                  "2b1c43fea817e9ad275ae546963");
-    TestChaCha20("0000000000000000000000000000000000000000000000000000000000000000", 0x0100000000000000ULL, 0,
+    TestChaCha20("", "0000000000000000000000000000000000000000000000000000000000000000", 0x0100000000000000ULL, 0,
                  "de9cba7bf3d69ef5e786dc63973f653a0b49e015adbff7134fcb7df137821031e85a050278a7084527214f73efc7fa5b52770"
                  "62eb7a0433e445f41e3");
-    TestChaCha20("0000000000000000000000000000000000000000000000000000000000000000", 1, 0,
+    TestChaCha20("", "0000000000000000000000000000000000000000000000000000000000000000", 1, 0,
                  "ef3fdfd6c61578fbf5cf35bd3dd33b8009631634d21e42ac33960bd138e50d32111e4caf237ee53ca8ad6426194a88545ddc4"
                  "97a0b466e7d6bbdb0041b2f586b");
-    TestChaCha20("000102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1f", 0x0706050403020100ULL, 0,
+    TestChaCha20("", "000102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1f", 0x0706050403020100ULL, 0,
                  "f798a189f195e66982105ffb640bb7757f579da31602fc93ec01ac56f85ac3c134a4547b733b46413042c9440049176905d3b"
                  "e59ea1c53f15916155c2be8241a38008b9a26bc35941e2444177c8ade6689de95264986d95889fb60e84629c9bd9a5acb1cc1"
                  "18be563eb9b3a4a472f82e09a7e778492b562ef7130e88dfe031c79db9d4f7c7a899151b9a475032b63fc385245fe054e3dd5"