Optimized siphash implementation

elichai · elichai · commit 409c2e345225 · 2022-10-12T16:55:47.000-04:00
diff --git a/src/crypto/siphash.cpp b/src/crypto/siphash.cpp
@@ -2,8 +2,11 @@
 // Distributed under the MIT software license, see the accompanying
 // file COPYING or http://www.opensource.org/licenses/mit-license.php.
 
+#include <crypto/common.h>
 #include <crypto/siphash.h>
 
+#include <algorithm>
+
 #define ROTL(x, b) (uint64_t)(((x) << (b)) | ((x) >> (64 - (b))))
 
 #define SIPROUND do { \
@@ -22,14 +25,14 @@ CSipHasher::CSipHasher(uint64_t k0, uint64_t k1)
     v[2] = 0x6c7967656e657261ULL ^ k0;
     v[3] = 0x7465646279746573ULL ^ k1;
     count = 0;
-    tmp = 0;
+    tail = 0;
 }
 
 CSipHasher& CSipHasher::Write(uint64_t data)
 {
     uint64_t v0 = v[0], v1 = v[1], v2 = v[2], v3 = v[3];
 
-    assert(count % 8 == 0);
+    assert((count & 0x07) == 0);
 
     v3 ^= data;
     SIPROUND;
@@ -45,30 +48,57 @@ CSipHasher& CSipHasher::Write(uint64_t data)
     return *this;
 }
 
+
+/// Load a uint64_t from 0 to 7 bytes.
+inline uint64_t ReadU64ByLenLE(const unsigned char* data, size_t len)
+{
+    assert(len < 8);
+    uint64_t out = 0;
+    for (size_t i = 0; i < len; ++i) {
+        out |= (uint64_t)data[i] << (i * 8);
+    }
+    return out;
+}
+
 CSipHasher& CSipHasher::Write(const unsigned char* data, size_t size)
 {
     uint64_t v0 = v[0], v1 = v[1], v2 = v[2], v3 = v[3];
-    uint64_t t = tmp;
-    uint8_t c = count;
-
-    while (size--) {
-        t |= ((uint64_t)(*(data++))) << (8 * (c % 8));
-        c++;
-        if ((c & 7) == 0) {
-            v3 ^= t;
+    auto ntail = count & 0x07;
+    count += size;
+
+    size_t needed = 0;
+
+    if (ntail != 0) {
+        needed = 8 - ntail;
+        tail |= ReadU64ByLenLE(data, std::min(size, needed)) << 8 * ntail;
+        if (size < needed) {
+            return *this;
+        } else {
+            v3 ^= tail;
             SIPROUND;
             SIPROUND;
-            v0 ^= t;
-            t = 0;
+            v0 ^= tail;
         }
     }
 
+    size_t len = size - needed;
+    auto left = len & 0x07;
+
+    auto i = needed;
+    while (i < len - left) {
+        uint64_t mi = ReadLE64(data + i);
+        v3 ^= mi;
+        SIPROUND;
+        SIPROUND;
+        v0 ^= mi;
+        i += 8;
+    }
+
     v[0] = v0;
     v[1] = v1;
     v[2] = v2;
     v[3] = v3;
-    count = c;
-    tmp = t;
+    tail = ReadU64ByLenLE(data + i, left);
 
     return *this;
 }
@@ -77,12 +107,14 @@ uint64_t CSipHasher::Finalize() const
 {
     uint64_t v0 = v[0], v1 = v[1], v2 = v[2], v3 = v[3];
 
-    uint64_t t = tmp | (((uint64_t)count) << 56);
+
+    uint64_t t = tail | (((uint64_t)count) << 56);
 
     v3 ^= t;
     SIPROUND;
     SIPROUND;
     v0 ^= t;
+
     v2 ^= 0xFF;
     SIPROUND;
     SIPROUND;
diff --git a/src/crypto/siphash.h b/src/crypto/siphash.h
@@ -14,8 +14,8 @@ class CSipHasher
 {
 private:
     uint64_t v[4];
-    uint64_t tmp;
-    uint8_t count; // Only the low 8 bits of the input size matter.
+    uint64_t tail; // bytes that weren't processed yet.
+    uint8_t count;  // total amount of bytes inputted.
 
 public:
     /** Construct a SipHash calculator initialized with 128-bit key (k0, k1) */
