Commit 00ff653576
Changed files (2)
lib
std
lib/std/hash/auto_hash.zig
@@ -343,7 +343,7 @@ test "testHash optional" {
const b: ?u32 = null;
try testing.expectEqual(testHash(a), testHash(@as(u32, 123)));
try testing.expect(testHash(a) != testHash(b));
- try testing.expectEqual(testHash(b), 0);
+ try testing.expectEqual(testHash(b), 0x409638ee2bde459); // wyhash empty input hash
}
test "testHash array" {
lib/std/hash/wyhash.zig
@@ -1,209 +1,240 @@
const std = @import("std");
-const mem = std.mem;
-
-const primes = [_]u64{
- 0xa0761d6478bd642f,
- 0xe7037ed1a0b428db,
- 0x8ebc6af09c88c6e3,
- 0x589965cc75374cc3,
- 0x1d8e4e27c47d124f,
-};
-fn read_bytes(comptime bytes: u8, data: []const u8) u64 {
- const T = std.meta.Int(.unsigned, 8 * bytes);
- return mem.readIntLittle(T, data[0..bytes]);
-}
+pub const Wyhash = struct {
+ const secret = [_]u64{
+ 0xa0761d6478bd642f,
+ 0xe7037ed1a0b428db,
+ 0x8ebc6af09c88c6e3,
+ 0x589965cc75374cc3,
+ };
+
+ a: u64,
+ b: u64,
+ state: [3]u64,
+ total_len: usize,
+
+ buf: [48]u8,
+ buf_len: usize,
-fn read_8bytes_swapped(data: []const u8) u64 {
- return (read_bytes(4, data) << 32 | read_bytes(4, data[4..]));
-}
+ pub fn init(seed: u64) Wyhash {
+ var self = Wyhash{
+ .a = undefined,
+ .b = undefined,
+ .state = undefined,
+ .total_len = 0,
+ .buf = undefined,
+ .buf_len = 0,
+ };
-fn mum(a: u64, b: u64) u64 {
- var r = std.math.mulWide(u64, a, b);
- r = (r >> 64) ^ r;
- return @truncate(u64, r);
-}
+ self.state[0] = seed ^ mix(seed ^ secret[0], secret[1]);
+ self.state[1] = self.state[0];
+ self.state[2] = self.state[0];
+ return self;
+ }
-fn mix0(a: u64, b: u64, seed: u64) u64 {
- return mum(a ^ seed ^ primes[0], b ^ seed ^ primes[1]);
-}
+ // This is subtly different from other hash function update calls. Wyhash requires the last
+ // full 48-byte block to be run through final1 if is exactly aligned to 48-bytes.
+ pub fn update(self: *Wyhash, input: []const u8) void {
+ self.total_len += input.len;
-fn mix1(a: u64, b: u64, seed: u64) u64 {
- return mum(a ^ seed ^ primes[2], b ^ seed ^ primes[3]);
-}
+ if (input.len <= 48 - self.buf_len) {
+ @memcpy(self.buf[self.buf_len..][0..input.len], input);
+ self.buf_len += input.len;
+ return;
+ }
-// Wyhash version which does not store internal state for handling partial buffers.
-// This is needed so that we can maximize the speed for the short key case, which will
-// use the non-iterative api which the public Wyhash exposes.
-const WyhashStateless = struct {
- seed: u64,
- msg_len: usize,
+ var i: usize = 0;
- pub fn init(seed: u64) WyhashStateless {
- return WyhashStateless{
- .seed = seed,
- .msg_len = 0,
- };
- }
+ if (self.buf_len > 0) {
+ i = 48 - self.buf_len;
+ @memcpy(self.buf[self.buf_len..][0..i], input[0..i]);
+ self.round(&self.buf);
+ self.buf_len = 0;
+ }
- fn round(self: *WyhashStateless, b: []const u8) void {
- std.debug.assert(b.len == 32);
+ while (i + 48 < input.len) : (i += 48) {
+ self.round(input[i..][0..48]);
+ }
- self.seed = mix0(
- read_bytes(8, b[0..]),
- read_bytes(8, b[8..]),
- self.seed,
- ) ^ mix1(
- read_bytes(8, b[16..]),
- read_bytes(8, b[24..]),
- self.seed,
- );
+ const remaining_bytes = input[i..];
+ @memcpy(self.buf[0..remaining_bytes.len], remaining_bytes);
+ self.buf_len = remaining_bytes.len;
}
- pub fn update(self: *WyhashStateless, b: []const u8) void {
- std.debug.assert(b.len % 32 == 0);
-
- var off: usize = 0;
- while (off < b.len) : (off += 32) {
- @call(.always_inline, round, .{ self, b[off..][0..32] });
+ pub fn final(self: *Wyhash) u64 {
+ var input = self.buf[0..self.buf_len];
+ var newSelf = self.shallowCopy(); // ensure idempotency
+
+ if (self.total_len <= 16) {
+ newSelf.smallKey(input);
+ } else {
+ if (self.buf_len < 16) {
+ var scratch: [16]u8 = undefined;
+ const rem = 16 - self.buf_len;
+ @memcpy(scratch[0..rem], self.buf[self.buf.len - rem ..][0..rem]);
+ @memcpy(scratch[rem..][0..self.buf_len], self.buf[0..self.buf_len]);
+
+ // Same as input with lookbehind to pad to 16-bytes
+ input = scratch[rem..];
+ }
+
+ newSelf.final0();
+ newSelf.final1(input);
}
- self.msg_len += b.len;
- }
-
- pub fn final(self: *WyhashStateless, b: []const u8) u64 {
- std.debug.assert(b.len < 32);
-
- const seed = self.seed;
- const rem_len = @intCast(u5, b.len);
- const rem_key = b[0..rem_len];
-
- self.seed = switch (rem_len) {
- 0 => seed,
- 1 => mix0(read_bytes(1, rem_key), primes[4], seed),
- 2 => mix0(read_bytes(2, rem_key), primes[4], seed),
- 3 => mix0((read_bytes(2, rem_key) << 8) | read_bytes(1, rem_key[2..]), primes[4], seed),
- 4 => mix0(read_bytes(4, rem_key), primes[4], seed),
- 5 => mix0((read_bytes(4, rem_key) << 8) | read_bytes(1, rem_key[4..]), primes[4], seed),
- 6 => mix0((read_bytes(4, rem_key) << 16) | read_bytes(2, rem_key[4..]), primes[4], seed),
- 7 => mix0((read_bytes(4, rem_key) << 24) | (read_bytes(2, rem_key[4..]) << 8) | read_bytes(1, rem_key[6..]), primes[4], seed),
- 8 => mix0(read_8bytes_swapped(rem_key), primes[4], seed),
- 9 => mix0(read_8bytes_swapped(rem_key), read_bytes(1, rem_key[8..]), seed),
- 10 => mix0(read_8bytes_swapped(rem_key), read_bytes(2, rem_key[8..]), seed),
- 11 => mix0(read_8bytes_swapped(rem_key), (read_bytes(2, rem_key[8..]) << 8) | read_bytes(1, rem_key[10..]), seed),
- 12 => mix0(read_8bytes_swapped(rem_key), read_bytes(4, rem_key[8..]), seed),
- 13 => mix0(read_8bytes_swapped(rem_key), (read_bytes(4, rem_key[8..]) << 8) | read_bytes(1, rem_key[12..]), seed),
- 14 => mix0(read_8bytes_swapped(rem_key), (read_bytes(4, rem_key[8..]) << 16) | read_bytes(2, rem_key[12..]), seed),
- 15 => mix0(read_8bytes_swapped(rem_key), (read_bytes(4, rem_key[8..]) << 24) | (read_bytes(2, rem_key[12..]) << 8) | read_bytes(1, rem_key[14..]), seed),
- 16 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed),
- 17 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1(read_bytes(1, rem_key[16..]), primes[4], seed),
- 18 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1(read_bytes(2, rem_key[16..]), primes[4], seed),
- 19 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1((read_bytes(2, rem_key[16..]) << 8) | read_bytes(1, rem_key[18..]), primes[4], seed),
- 20 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1(read_bytes(4, rem_key[16..]), primes[4], seed),
- 21 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1((read_bytes(4, rem_key[16..]) << 8) | read_bytes(1, rem_key[20..]), primes[4], seed),
- 22 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1((read_bytes(4, rem_key[16..]) << 16) | read_bytes(2, rem_key[20..]), primes[4], seed),
- 23 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1((read_bytes(4, rem_key[16..]) << 24) | (read_bytes(2, rem_key[20..]) << 8) | read_bytes(1, rem_key[22..]), primes[4], seed),
- 24 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1(read_8bytes_swapped(rem_key[16..]), primes[4], seed),
- 25 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1(read_8bytes_swapped(rem_key[16..]), read_bytes(1, rem_key[24..]), seed),
- 26 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1(read_8bytes_swapped(rem_key[16..]), read_bytes(2, rem_key[24..]), seed),
- 27 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1(read_8bytes_swapped(rem_key[16..]), (read_bytes(2, rem_key[24..]) << 8) | read_bytes(1, rem_key[26..]), seed),
- 28 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1(read_8bytes_swapped(rem_key[16..]), read_bytes(4, rem_key[24..]), seed),
- 29 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1(read_8bytes_swapped(rem_key[16..]), (read_bytes(4, rem_key[24..]) << 8) | read_bytes(1, rem_key[28..]), seed),
- 30 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1(read_8bytes_swapped(rem_key[16..]), (read_bytes(4, rem_key[24..]) << 16) | read_bytes(2, rem_key[28..]), seed),
- 31 => mix0(read_8bytes_swapped(rem_key), read_8bytes_swapped(rem_key[8..]), seed) ^ mix1(read_8bytes_swapped(rem_key[16..]), (read_bytes(4, rem_key[24..]) << 24) | (read_bytes(2, rem_key[28..]) << 8) | read_bytes(1, rem_key[30..]), seed),
+ return newSelf.final2();
+ }
+
+ // Copies the core wyhash state but not any internal buffers.
+ inline fn shallowCopy(self: *Wyhash) Wyhash {
+ return .{
+ .a = self.a,
+ .b = self.b,
+ .state = self.state,
+ .total_len = self.total_len,
+ .buf = undefined,
+ .buf_len = undefined,
};
+ }
- self.msg_len += b.len;
- return mum(self.seed ^ self.msg_len, primes[4]);
+ inline fn smallKey(self: *Wyhash, input: []const u8) void {
+ std.debug.assert(input.len <= 16);
+
+ if (input.len >= 4) {
+ const end = input.len - 4;
+ const quarter = (input.len >> 3) << 2;
+ self.a = (read(4, input[0..]) << 32) | read(4, input[quarter..]);
+ self.b = (read(4, input[end..]) << 32) | read(4, input[end - quarter ..]);
+ } else if (input.len > 0) {
+ self.a = (@as(u64, input[0]) << 16) | (@as(u64, input[input.len >> 1]) << 8) | input[input.len - 1];
+ self.b = 0;
+ } else {
+ self.a = 0;
+ self.b = 0;
+ }
}
- pub fn hash(seed: u64, input: []const u8) u64 {
- const aligned_len = input.len - (input.len % 32);
+ inline fn round(self: *Wyhash, input: *const [48]u8) void {
+ inline for (0..3) |i| {
+ const a = read(8, input[8 * (2 * i) ..]);
+ const b = read(8, input[8 * (2 * i + 1) ..]);
+ self.state[i] = mix(a ^ secret[i + 1], b ^ self.state[i]);
+ }
+ }
- var c = WyhashStateless.init(seed);
- @call(.always_inline, update, .{ &c, input[0..aligned_len] });
- return @call(.always_inline, final, .{ &c, input[aligned_len..] });
+ inline fn read(comptime bytes: usize, data: []const u8) u64 {
+ std.debug.assert(bytes <= 8);
+ const T = std.meta.Int(.unsigned, 8 * bytes);
+ return @as(u64, std.mem.readIntLittle(T, data[0..bytes]));
}
-};
-/// Fast non-cryptographic 64bit hash function.
-/// See https://github.com/wangyi-fudan/wyhash
-pub const Wyhash = struct {
- state: WyhashStateless,
+ inline fn mum(a: *u64, b: *u64) void {
+ const x = @as(u128, a.*) *% b.*;
+ a.* = @truncate(u64, x);
+ b.* = @truncate(u64, x >> 64);
+ }
- buf: [32]u8,
- buf_len: usize,
+ inline fn mix(a_: u64, b_: u64) u64 {
+ var a = a_;
+ var b = b_;
+ mum(&a, &b);
+ return a ^ b;
+ }
- pub fn init(seed: u64) Wyhash {
- return Wyhash{
- .state = WyhashStateless.init(seed),
- .buf = undefined,
- .buf_len = 0,
- };
+ inline fn final0(self: *Wyhash) void {
+ self.state[0] ^= self.state[1] ^ self.state[2];
}
- pub fn update(self: *Wyhash, b: []const u8) void {
- var off: usize = 0;
+ // Input must reside in a 16-byte buffer. The input slice passed be offset into it in which
+ // case this function will index in front of the slice.
+ inline fn final1(self: *Wyhash, input: []const u8) void {
+ std.debug.assert(input.len <= 48);
- if (self.buf_len != 0 and self.buf_len + b.len >= 32) {
- off += 32 - self.buf_len;
- @memcpy(self.buf[self.buf_len..][0..off], b[0..off]);
- self.state.update(self.buf[0..]);
- self.buf_len = 0;
+ var i: usize = 0;
+ while (i + 16 < input.len) : (i += 16) {
+ self.state[0] = mix(read(8, input[i..]) ^ secret[1], read(8, input[i + 8 ..]) ^ self.state[0]);
}
- const remain_len = b.len - off;
- const aligned_len = remain_len - (remain_len % 32);
- self.state.update(b[off .. off + aligned_len]);
-
- const src = b[off + aligned_len ..];
- @memcpy(self.buf[self.buf_len..][0..src.len], src);
- self.buf_len += @intCast(u8, b[off + aligned_len ..].len);
+ // Possible lookbehind past pointer start.
+ self.a = read(8, (input.ptr + input.len - 16)[0..8]);
+ self.b = read(8, (input.ptr + input.len - 8)[0..8]);
}
- pub fn final(self: *Wyhash) u64 {
- const rem_key = self.buf[0..self.buf_len];
-
- return self.state.final(rem_key);
+ inline fn final2(self: *Wyhash) u64 {
+ self.a ^= secret[1];
+ self.b ^= self.state[0];
+ mum(&self.a, &self.b);
+ return mix(self.a ^ secret[0] ^ self.total_len, self.b ^ secret[1]);
}
pub fn hash(seed: u64, input: []const u8) u64 {
- return WyhashStateless.hash(seed, input);
+ var self = Wyhash.init(seed);
+
+ if (input.len <= 16) {
+ self.smallKey(input);
+ } else {
+ var i: usize = 0;
+ if (input.len >= 48) {
+ while (i + 48 < input.len) : (i += 48) {
+ self.round(input[i..][0..48]);
+ }
+ self.final0();
+ }
+ self.final1(input[i..]);
+ }
+
+ self.total_len = input.len;
+ return self.final2();
}
};
const expectEqual = std.testing.expectEqual;
+const TestVector = struct {
+ expected: u64,
+ seed: u64,
+ input: []const u8,
+};
+
+// Run https://github.com/wangyi-fudan/wyhash/blob/77e50f267fbc7b8e2d09f2d455219adb70ad4749/test_vector.cpp directly.
+const vectors = [_]TestVector{
+ .{ .seed = 0, .expected = 0x409638ee2bde459, .input = "" },
+ .{ .seed = 1, .expected = 0xa8412d091b5fe0a9, .input = "a" },
+ .{ .seed = 2, .expected = 0x32dd92e4b2915153, .input = "abc" },
+ .{ .seed = 3, .expected = 0x8619124089a3a16b, .input = "message digest" },
+ .{ .seed = 4, .expected = 0x7a43afb61d7f5f40, .input = "abcdefghijklmnopqrstuvwxyz" },
+ .{ .seed = 5, .expected = 0xff42329b90e50d58, .input = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789" },
+ .{ .seed = 6, .expected = 0xc39cab13b115aad3, .input = "12345678901234567890123456789012345678901234567890123456789012345678901234567890" },
+};
+
test "test vectors" {
- const hash = Wyhash.hash;
-
- try expectEqual(hash(0, ""), 0x0);
- try expectEqual(hash(1, "a"), 0xbed235177f41d328);
- try expectEqual(hash(2, "abc"), 0xbe348debe59b27c3);
- try expectEqual(hash(3, "message digest"), 0x37320f657213a290);
- try expectEqual(hash(4, "abcdefghijklmnopqrstuvwxyz"), 0xd0b270e1d8a7019c);
- try expectEqual(hash(5, "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789"), 0x602a1894d3bbfe7f);
- try expectEqual(hash(6, "12345678901234567890123456789012345678901234567890123456789012345678901234567890"), 0x829e9c148b75970e);
+ for (vectors) |e| {
+ try expectEqual(e.expected, Wyhash.hash(e.seed, e.input));
+ }
}
test "test vectors streaming" {
- var wh = Wyhash.init(5);
- for ("ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789") |e| {
- wh.update(mem.asBytes(&e));
+ const step = 5;
+
+ for (vectors) |e| {
+ var wh = Wyhash.init(e.seed);
+ var i: usize = 0;
+ while (i < e.input.len) : (i += step) {
+ const len = if (i + step > e.input.len) e.input.len - i else step;
+ wh.update(e.input[i..][0..len]);
+ }
+ try expectEqual(e.expected, wh.final());
}
- try expectEqual(wh.final(), 0x602a1894d3bbfe7f);
+}
- const pattern = "1234567890";
- const count = 8;
- const result = 0x829e9c148b75970e;
- try expectEqual(Wyhash.hash(6, pattern ** 8), result);
+test "test ensure idempotent final call" {
+ const e: TestVector = .{ .seed = 6, .expected = 0xc39cab13b115aad3, .input = "12345678901234567890123456789012345678901234567890123456789012345678901234567890" };
+ var wh = Wyhash.init(e.seed);
+ wh.update(e.input);
- wh = Wyhash.init(6);
- var i: u32 = 0;
- while (i < count) : (i += 1) {
- wh.update(pattern);
+ for (0..10) |_| {
+ try expectEqual(e.expected, wh.final());
}
- try expectEqual(wh.final(), result);
}
test "iterative non-divisible update" {