diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index c19ef48829..2ee0e85ccf 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -25,6 +25,7 @@ Here are some examples: * [Iterative Replacement of C with Zig](http://tiehuis.github.io/blog/zig1.html) * [The Right Tool for the Right Job: Redis Modules & Zig](https://www.youtube.com/watch?v=eCHM8-_poZY) + * [Writing a small ray tracer in Rust and Zig](https://nelari.us/post/raytracer_with_rust_and_zig/) Zig is a brand new language, with no advertising budget. Word of mouth is the only way people find out about the project, and the more people hear about it, @@ -45,8 +46,8 @@ The most highly regarded argument in such a discussion is a real world use case. The issue label [Contributor Friendly](https://github.com/ziglang/zig/issues?q=is%3Aissue+is%3Aopen+label%3A%22contributor+friendly%22) -exists to help contributors find issues that are "limited in scope and/or -knowledge of Zig internals." +exists to help you find issues that are **limited in scope and/or +knowledge of Zig internals.** ### Editing Source Code @@ -61,8 +62,7 @@ To test changes, do the following from the build directory: 1. Run `make install` (on POSIX) or `msbuild -p:Configuration=Release INSTALL.vcxproj` (on Windows). -2. `bin/zig build --build-file ../build.zig test` (on POSIX) or - `bin\zig.exe build --build-file ..\build.zig test` (on Windows). +2. `bin/zig build test` (on POSIX) or `bin\zig.exe build test` (on Windows). That runs the whole test suite, which does a lot of extra testing that you likely won't always need, and can take upwards of 2 hours. This is what the @@ -79,8 +79,8 @@ Another example is choosing a different set of things to test. For example, not the other ones. Combining this suggestion with the previous one, you could do this: -`bin/zig build --build-file ../build.zig test-std -Dskip-release` (on POSIX) or -`bin\zig.exe build --build-file ..\build.zig test-std -Dskip-release` (on Windows). +`bin/zig build test-std -Dskip-release` (on POSIX) or +`bin\zig.exe build test-std -Dskip-release` (on Windows). This will run only the standard library tests, in debug mode only, for all targets (it will cross-compile the tests for non-native targets but not run diff --git a/doc/langref.html.in b/doc/langref.html.in index b5fe464c35..ac381e00b2 100644 --- a/doc/langref.html.in +++ b/doc/langref.html.in @@ -6330,6 +6330,22 @@ comptime { TODO right now bool is not accepted. Also I think we could make non powers of 2 work fine, maybe we can remove this restriction

+

+ Supported operations: +

+ {#header_close#} {#header_open|@bitCast#}
{#syntax#}@bitCast(comptime DestType: type, value: var) DestType{#endsyntax#}
diff --git a/std/build.zig b/std/build.zig index 6be88ae7ec..9393d72c15 100644 --- a/std/build.zig +++ b/std/build.zig @@ -805,11 +805,7 @@ pub const Builder = struct { return name; } const full_path = try fs.path.join(self.allocator, [_][]const u8{ search_prefix, "bin", self.fmt("{}{}", name, exe_extension) }); - if (fs.path.real(self.allocator, full_path)) |real_path| { - return real_path; - } else |_| { - continue; - } + return fs.realpathAlloc(self.allocator, full_path) catch continue; } } if (self.env_map.get("PATH")) |PATH| { @@ -817,14 +813,10 @@ pub const Builder = struct { if (fs.path.isAbsolute(name)) { return name; } - var it = mem.tokenize(PATH, []u8{fs.path.delimiter}); + var it = mem.tokenize(PATH, [_]u8{fs.path.delimiter}); while (it.next()) |path| { const full_path = try fs.path.join(self.allocator, [_][]const u8{ path, self.fmt("{}{}", name, exe_extension) }); - if (fs.path.real(self.allocator, full_path)) |real_path| { - return real_path; - } else |_| { - continue; - } + return fs.realpathAlloc(self.allocator, full_path) catch continue; } } } @@ -834,11 +826,7 @@ pub const Builder = struct { } for (paths) |path| { const full_path = try fs.path.join(self.allocator, [_][]const u8{ path, self.fmt("{}{}", name, exe_extension) }); - if (fs.path.real(self.allocator, full_path)) |real_path| { - return real_path; - } else |_| { - continue; - } + return fs.realpathAlloc(self.allocator, full_path) catch continue; } } return error.FileNotFound; @@ -904,6 +892,15 @@ pub const Builder = struct { } }; +test "builder.findProgram compiles" { + //allocator: *Allocator, + //zig_exe: []const u8, + //build_root: []const u8, + //cache_root: []const u8, + const builder = try Builder.create(std.heap.direct_allocator, "zig", "zig-cache", "zig-cache"); + _ = builder.findProgram([_][]const u8{}, [_][]const u8{}) catch null; +} + pub const Version = struct { major: u32, minor: u32, @@ -1122,6 +1119,9 @@ pub const Target = union(enum) { } pub fn libPrefix(self: Target) []const u8 { + if (self.isWasm()) { + return ""; + } switch (self.getAbi()) { .msvc => return "", else => return "lib", diff --git a/std/debug.zig b/std/debug.zig index 32f96d3e15..d1c17343ef 100644 --- a/std/debug.zig +++ b/std/debug.zig @@ -1024,8 +1024,7 @@ pub fn openElfDebugInfo( elf_seekable_stream: *DwarfSeekableStream, elf_in_stream: *DwarfInStream, ) !DwarfInfo { - var efile: elf.Elf = undefined; - try efile.openStream(allocator, elf_seekable_stream, elf_in_stream); + var efile = try elf.Elf.openStream(allocator, elf_seekable_stream, elf_in_stream); errdefer efile.close(); var di = DwarfInfo{ diff --git a/std/elf.zig b/std/elf.zig index c605a177a5..37635895fd 100644 --- a/std/elf.zig +++ b/std/elf.zig @@ -356,7 +356,6 @@ pub const SectionHeader = struct { pub const Elf = struct { seekable_stream: *io.SeekableStream(anyerror, anyerror), in_stream: *io.InStream(anyerror), - auto_close_stream: bool, is_64: bool, endian: builtin.Endian, file_type: FileType, @@ -368,25 +367,23 @@ pub const Elf = struct { string_section: *SectionHeader, section_headers: []SectionHeader, allocator: *mem.Allocator, - prealloc_file: File, /// Call close when done. - pub fn openPath(elf: *Elf, allocator: *mem.Allocator, path: []const u8) !void { + pub fn openPath(allocator: *mem.Allocator, path: []const u8) !Elf { @compileError("TODO implement"); } /// Call close when done. - pub fn openFile(elf: *Elf, allocator: *mem.Allocator, file: File) !void { + pub fn openFile(allocator: *mem.Allocator, file: File) !Elf { @compileError("TODO implement"); } pub fn openStream( - elf: *Elf, allocator: *mem.Allocator, seekable_stream: *io.SeekableStream(anyerror, anyerror), in: *io.InStream(anyerror), - ) !void { - elf.auto_close_stream = false; + ) !Elf { + var elf: Elf = undefined; elf.allocator = allocator; elf.seekable_stream = seekable_stream; elf.in_stream = in; @@ -523,12 +520,12 @@ pub const Elf = struct { // not a string table return error.InvalidFormat; } + + return elf; } pub fn close(elf: *Elf) void { elf.allocator.free(elf.section_headers); - - if (elf.auto_close_stream) elf.prealloc_file.close(); } pub fn findSection(elf: *Elf, name: []const u8) !?*SectionHeader { diff --git a/std/hash.zig b/std/hash.zig index 148504aa39..648f34b11d 100644 --- a/std/hash.zig +++ b/std/hash.zig @@ -1,6 +1,9 @@ const adler = @import("hash/adler.zig"); pub const Adler32 = adler.Adler32; +const auto_hash = @import("hash/auto_hash.zig"); +pub const autoHash = auto_hash.autoHash; + // pub for polynomials + generic crc32 construction pub const crc = @import("hash/crc.zig"); pub const Crc32 = crc.Crc32; @@ -16,6 +19,8 @@ pub const SipHash128 = siphash.SipHash128; pub const murmur = @import("hash/murmur.zig"); pub const Murmur2_32 = murmur.Murmur2_32; + + pub const Murmur2_64 = murmur.Murmur2_64; pub const Murmur3_32 = murmur.Murmur3_32; @@ -23,11 +28,16 @@ pub const cityhash = @import("hash/cityhash.zig"); pub const CityHash32 = cityhash.CityHash32; pub const CityHash64 = cityhash.CityHash64; +const wyhash = @import("hash/wyhash.zig"); +pub const Wyhash = wyhash.Wyhash; + test "hash" { _ = @import("hash/adler.zig"); + _ = @import("hash/auto_hash.zig"); _ = @import("hash/crc.zig"); _ = @import("hash/fnv.zig"); _ = @import("hash/siphash.zig"); _ = @import("hash/murmur.zig"); _ = @import("hash/cityhash.zig"); + _ = @import("hash/wyhash.zig"); } diff --git a/std/hash/auto_hash.zig b/std/hash/auto_hash.zig new file mode 100644 index 0000000000..251f7e798a --- /dev/null +++ b/std/hash/auto_hash.zig @@ -0,0 +1,211 @@ +const std = @import("std"); +const builtin = @import("builtin"); +const mem = std.mem; +const meta = std.meta; + +/// Provides generic hashing for any eligible type. +/// Only hashes `key` itself, pointers are not followed. +pub fn autoHash(hasher: var, key: var) void { + const Key = @typeOf(key); + switch (@typeInfo(Key)) { + .NoReturn, + .Opaque, + .Undefined, + .ArgTuple, + .Void, + .Null, + .BoundFn, + .ComptimeFloat, + .ComptimeInt, + .Type, + .EnumLiteral, + .Frame, + => @compileError("cannot hash this type"), + + // Help the optimizer see that hashing an int is easy by inlining! + // TODO Check if the situation is better after #561 is resolved. + .Int => @inlineCall(hasher.update, std.mem.asBytes(&key)), + + .Float => |info| autoHash(hasher, @bitCast(@IntType(false, info.bits), key)), + + .Bool => autoHash(hasher, @boolToInt(key)), + .Enum => autoHash(hasher, @enumToInt(key)), + .ErrorSet => autoHash(hasher, @errorToInt(key)), + .AnyFrame, .Fn => autoHash(hasher, @ptrToInt(key)), + + .Pointer => |info| switch (info.size) { + builtin.TypeInfo.Pointer.Size.One, + builtin.TypeInfo.Pointer.Size.Many, + builtin.TypeInfo.Pointer.Size.C, + => autoHash(hasher, @ptrToInt(key)), + + builtin.TypeInfo.Pointer.Size.Slice => { + autoHash(hasher, key.ptr); + autoHash(hasher, key.len); + }, + }, + + .Optional => if (key) |k| autoHash(hasher, k), + + .Array => { + // TODO detect via a trait when Key has no padding bits to + // hash it as an array of bytes. + // Otherwise, hash every element. + for (key) |element| { + autoHash(hasher, element); + } + }, + + .Vector => |info| { + if (info.child.bit_count % 8 == 0) { + // If there's no unused bits in the child type, we can just hash + // this as an array of bytes. + hasher.update(mem.asBytes(&key)); + } else { + // Otherwise, hash every element. + // TODO remove the copy to an array once field access is done. + const array: [info.len]info.child = key; + comptime var i: u32 = 0; + inline while (i < info.len) : (i += 1) { + autoHash(hasher, array[i]); + } + } + }, + + .Struct => |info| { + // TODO detect via a trait when Key has no padding bits to + // hash it as an array of bytes. + // Otherwise, hash every field. + inline for (info.fields) |field| { + // We reuse the hash of the previous field as the seed for the + // next one so that they're dependant. + autoHash(hasher, @field(key, field.name)); + } + }, + + .Union => |info| blk: { + if (info.tag_type) |tag_type| { + const tag = meta.activeTag(key); + const s = autoHash(hasher, tag); + inline for (info.fields) |field| { + const enum_field = field.enum_field.?; + if (enum_field.value == @enumToInt(tag)) { + autoHash(hasher, @field(key, enum_field.name)); + // TODO use a labelled break when it does not crash the compiler. + // break :blk; + return; + } + } + unreachable; + } else @compileError("cannot hash untagged union type: " ++ @typeName(Key) ++ ", provide your own hash function"); + }, + + .ErrorUnion => blk: { + const payload = key catch |err| { + autoHash(hasher, err); + break :blk; + }; + autoHash(hasher, payload); + }, + } +} + +const testing = std.testing; +const Wyhash = std.hash.Wyhash; + +fn testAutoHash(key: var) u64 { + // Any hash could be used here, for testing autoHash. + var hasher = Wyhash.init(0); + autoHash(&hasher, key); + return hasher.final(); +} + +test "autoHash slice" { + // Allocate one array dynamically so that we're assured it is not merged + // with the other by the optimization passes. + const array1 = try std.heap.direct_allocator.create([6]u32); + defer std.heap.direct_allocator.destroy(array1); + array1.* = [_]u32{ 1, 2, 3, 4, 5, 6 }; + const array2 = [_]u32{ 1, 2, 3, 4, 5, 6 }; + const a = array1[0..]; + const b = array2[0..]; + const c = array1[0..3]; + testing.expect(testAutoHash(a) == testAutoHash(a)); + testing.expect(testAutoHash(a) != testAutoHash(array1)); + testing.expect(testAutoHash(a) != testAutoHash(b)); + testing.expect(testAutoHash(a) != testAutoHash(c)); +} + +test "testAutoHash optional" { + const a: ?u32 = 123; + const b: ?u32 = null; + testing.expectEqual(testAutoHash(a), testAutoHash(u32(123))); + testing.expect(testAutoHash(a) != testAutoHash(b)); + testing.expectEqual(testAutoHash(b), 0); +} + +test "testAutoHash array" { + const a = [_]u32{ 1, 2, 3 }; + const h = testAutoHash(a); + var hasher = Wyhash.init(0); + autoHash(&hasher, u32(1)); + autoHash(&hasher, u32(2)); + autoHash(&hasher, u32(3)); + testing.expectEqual(h, hasher.final()); +} + +test "testAutoHash struct" { + const Foo = struct { + a: u32 = 1, + b: u32 = 2, + c: u32 = 3, + }; + const f = Foo{}; + const h = testAutoHash(f); + var hasher = Wyhash.init(0); + autoHash(&hasher, u32(1)); + autoHash(&hasher, u32(2)); + autoHash(&hasher, u32(3)); + testing.expectEqual(h, hasher.final()); +} + +test "testAutoHash union" { + const Foo = union(enum) { + A: u32, + B: f32, + C: u32, + }; + + const a = Foo{ .A = 18 }; + var b = Foo{ .B = 12.34 }; + const c = Foo{ .C = 18 }; + testing.expect(testAutoHash(a) == testAutoHash(a)); + testing.expect(testAutoHash(a) != testAutoHash(b)); + testing.expect(testAutoHash(a) != testAutoHash(c)); + + b = Foo{ .A = 18 }; + testing.expect(testAutoHash(a) == testAutoHash(b)); +} + +test "testAutoHash vector" { + const a: @Vector(4, u32) = [_]u32{ 1, 2, 3, 4 }; + const b: @Vector(4, u32) = [_]u32{ 1, 2, 3, 5 }; + const c: @Vector(4, u31) = [_]u31{ 1, 2, 3, 4 }; + testing.expect(testAutoHash(a) == testAutoHash(a)); + testing.expect(testAutoHash(a) != testAutoHash(b)); + testing.expect(testAutoHash(a) != testAutoHash(c)); +} + +test "testAutoHash error union" { + const Errors = error{Test}; + const Foo = struct { + a: u32 = 1, + b: u32 = 2, + c: u32 = 3, + }; + const f = Foo{}; + const g: Errors!Foo = Errors.Test; + testing.expect(testAutoHash(f) != testAutoHash(g)); + testing.expect(testAutoHash(f) == testAutoHash(Foo{})); + testing.expect(testAutoHash(g) == testAutoHash(Errors.Test)); +} diff --git a/std/hash/throughput_test.zig b/std/hash/throughput_test.zig new file mode 100644 index 0000000000..4b7e8ef344 --- /dev/null +++ b/std/hash/throughput_test.zig @@ -0,0 +1,148 @@ +const builtin = @import("builtin"); +const std = @import("std"); +const time = std.time; +const Timer = time.Timer; +const hash = std.hash; + +const KiB = 1024; +const MiB = 1024 * KiB; +const GiB = 1024 * MiB; + +var prng = std.rand.DefaultPrng.init(0); + +const Hash = struct { + ty: type, + name: []const u8, + init_u8s: ?[]const u8 = null, + init_u64: ?u64 = null, +}; + +const siphash_key = "0123456789abcdef"; + +const hashes = [_]Hash{ + Hash{ .ty = hash.Wyhash, .name = "wyhash", .init_u64 = 0 }, + Hash{ .ty = hash.SipHash64(1, 3), .name = "siphash(1,3)", .init_u8s = siphash_key }, + Hash{ .ty = hash.SipHash64(2, 4), .name = "siphash(2,4)", .init_u8s = siphash_key }, + Hash{ .ty = hash.Fnv1a_64, .name = "fnv1a" }, + Hash{ .ty = hash.Crc32, .name = "crc32" }, +}; + +const Result = struct { + hash: u64, + throughput: u64, +}; + +pub fn benchmarkHash(comptime H: var, bytes: usize) !Result { + var h = blk: { + if (H.init_u8s) |init| { + break :blk H.ty.init(init); + } + if (H.init_u64) |init| { + break :blk H.ty.init(init); + } + break :blk H.ty.init(); + }; + + var block: [8192]u8 = undefined; + prng.random.bytes(block[0..]); + + var offset: usize = 0; + var timer = try Timer.start(); + const start = timer.lap(); + while (offset < bytes) : (offset += block.len) { + h.update(block[0..]); + } + const end = timer.read(); + + const elapsed_s = @intToFloat(f64, end - start) / time.ns_per_s; + const throughput = @floatToInt(u64, @intToFloat(f64, bytes) / elapsed_s); + + return Result{ + .hash = h.final(), + .throughput = throughput, + }; +} + +fn usage() void { + std.debug.warn( + \\throughput_test [options] + \\ + \\Options: + \\ --filter [test-name] + \\ --seed [int] + \\ --count [int] + \\ --help + \\ + ); +} + +fn mode(comptime x: comptime_int) comptime_int { + return if (builtin.mode == builtin.Mode.Debug) x / 64 else x; +} + +// TODO(#1358): Replace with builtin formatted padding when available. +fn printPad(stdout: var, s: []const u8) !void { + var i: usize = 0; + while (i < 12 - s.len) : (i += 1) { + try stdout.print(" "); + } + try stdout.print("{}", s); +} + +pub fn main() !void { + var stdout_file = try std.io.getStdOut(); + var stdout_out_stream = stdout_file.outStream(); + const stdout = &stdout_out_stream.stream; + + var buffer: [1024]u8 = undefined; + var fixed = std.heap.FixedBufferAllocator.init(buffer[0..]); + const args = try std.process.argsAlloc(&fixed.allocator); + + var filter: ?[]u8 = ""; + var count: usize = mode(128 * MiB); + + var i: usize = 1; + while (i < args.len) : (i += 1) { + if (std.mem.eql(u8, args[i], "--seed")) { + i += 1; + if (i == args.len) { + usage(); + std.os.exit(1); + } + + const seed = try std.fmt.parseUnsigned(u32, args[i], 10); + prng.seed(seed); + } else if (std.mem.eql(u8, args[i], "--filter")) { + i += 1; + if (i == args.len) { + usage(); + std.os.exit(1); + } + + filter = args[i]; + } else if (std.mem.eql(u8, args[i], "--count")) { + i += 1; + if (i == args.len) { + usage(); + std.os.exit(1); + } + + const c = try std.fmt.parseUnsigned(u32, args[i], 10); + count = c * MiB; + } else if (std.mem.eql(u8, args[i], "--help")) { + usage(); + return; + } else { + usage(); + std.os.exit(1); + } + } + + inline for (hashes) |H| { + if (filter == null or std.mem.indexOf(u8, H.name, filter.?) != null) { + const result = try benchmarkHash(H, count); + try printPad(stdout, H.name); + try stdout.print(": {:4} MiB/s [{:16}]\n", result.throughput / (1 * MiB), result.hash); + } + } +} diff --git a/std/hash/wyhash.zig b/std/hash/wyhash.zig new file mode 100644 index 0000000000..dfa5156cad --- /dev/null +++ b/std/hash/wyhash.zig @@ -0,0 +1,135 @@ +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 { + return mem.readVarInt(u64, data[0..bytes], .Little); +} + +fn read_8bytes_swapped(data: []const u8) u64 { + return (read_bytes(4, data) << 32 | read_bytes(4, data[4..])); +} + +fn mum(a: u64, b: u64) u64 { + var r = std.math.mulWide(u64, a, b); + r = (r >> 64) ^ r; + return @truncate(u64, r); +} + +fn mix0(a: u64, b: u64, seed: u64) u64 { + return mum(a ^ seed ^ primes[0], b ^ seed ^ primes[1]); +} + +fn mix1(a: u64, b: u64, seed: u64) u64 { + return mum(a ^ seed ^ primes[2], b ^ seed ^ primes[3]); +} + +pub const Wyhash = struct { + seed: u64, + msg_len: usize, + + pub fn init(seed: u64) Wyhash { + return Wyhash{ + .seed = seed, + .msg_len = 0, + }; + } + + fn round(self: *Wyhash, b: []const u8) void { + std.debug.assert(b.len == 32); + + 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, + ); + } + + fn partial(self: *Wyhash, b: []const u8) void { + const rem_key = b; + const rem_len = b.len; + + var seed = self.seed; + seed = switch (@intCast(u5, 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), + }; + self.seed = seed; + } + + pub fn update(self: *Wyhash, b: []const u8) void { + var off: usize = 0; + + // Full middle blocks. + while (off + 32 <= b.len) : (off += 32) { + @inlineCall(self.round, b[off .. off + 32]); + } + + self.partial(b[off..]); + self.msg_len += b.len; + } + + pub fn final(self: *Wyhash) u64 { + return mum(self.seed ^ self.msg_len, primes[4]); + } + + pub fn hash(seed: u64, input: []const u8) u64 { + var c = Wyhash.init(seed); + c.update(input); + return c.final(); + } +}; + +test "test vectors" { + const expectEqual = std.testing.expectEqual; + const hash = Wyhash.hash; + + expectEqual(hash(0, ""), 0x0); + expectEqual(hash(1, "a"), 0xbed235177f41d328); + expectEqual(hash(2, "abc"), 0xbe348debe59b27c3); + expectEqual(hash(3, "message digest"), 0x37320f657213a290); + expectEqual(hash(4, "abcdefghijklmnopqrstuvwxyz"), 0xd0b270e1d8a7019c); + expectEqual(hash(5, "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789"), 0x602a1894d3bbfe7f); + expectEqual(hash(6, "12345678901234567890123456789012345678901234567890123456789012345678901234567890"), 0x829e9c148b75970e); +} diff --git a/std/hash_map.zig b/std/hash_map.zig index 431fbb35ab..ab3c4c248d 100644 --- a/std/hash_map.zig +++ b/std/hash_map.zig @@ -4,6 +4,9 @@ const assert = debug.assert; const testing = std.testing; const math = std.math; const mem = std.mem; +const meta = std.meta; +const autoHash = std.hash.autoHash; +const Wyhash = std.hash.Wyhash; const Allocator = mem.Allocator; const builtin = @import("builtin"); @@ -448,15 +451,17 @@ test "iterator hash map" { try reset_map.putNoClobber(2, 22); try reset_map.putNoClobber(3, 33); + // TODO this test depends on the hashing algorithm, because it assumes the + // order of the elements in the hashmap. This should not be the case. var keys = [_]i32{ + 1, 3, 2, - 1, }; var values = [_]i32{ + 11, 33, 22, - 11, }; var it = reset_map.iterator(); @@ -518,8 +523,9 @@ pub fn getTrivialEqlFn(comptime K: type) (fn (K, K) bool) { pub fn getAutoHashFn(comptime K: type) (fn (K) u32) { return struct { fn hash(key: K) u32 { - comptime var rng = comptime std.rand.DefaultPrng.init(0); - return autoHash(key, &rng.random, u32); + var hasher = Wyhash.init(0); + autoHash(&hasher, key); + return @truncate(u32, hasher.final()); } }.hash; } @@ -527,116 +533,7 @@ pub fn getAutoHashFn(comptime K: type) (fn (K) u32) { pub fn getAutoEqlFn(comptime K: type) (fn (K, K) bool) { return struct { fn eql(a: K, b: K) bool { - return autoEql(a, b); + return meta.eql(a, b); } }.eql; } - -// TODO improve these hash functions -pub fn autoHash(key: var, comptime rng: *std.rand.Random, comptime HashInt: type) HashInt { - switch (@typeInfo(@typeOf(key))) { - .NoReturn, - .Opaque, - .Undefined, - .ArgTuple, - .Frame, - .AnyFrame, - => @compileError("cannot hash this type"), - - .Void, - .Null, - => return 0, - - .Int => |info| { - const unsigned_x = @bitCast(@IntType(false, info.bits), key); - if (info.bits <= HashInt.bit_count) { - return HashInt(unsigned_x) ^ comptime rng.scalar(HashInt); - } else { - return @truncate(HashInt, unsigned_x ^ comptime rng.scalar(@typeOf(unsigned_x))); - } - }, - - .Float => |info| { - return autoHash(@bitCast(@IntType(false, info.bits), key), rng, HashInt); - }, - .Bool => return autoHash(@boolToInt(key), rng, HashInt), - .Enum => return autoHash(@enumToInt(key), rng, HashInt), - .ErrorSet => return autoHash(@errorToInt(key), rng, HashInt), - .Fn => return autoHash(@ptrToInt(key), rng, HashInt), - - .BoundFn, - .ComptimeFloat, - .ComptimeInt, - .Type, - .EnumLiteral, - => return 0, - - .Pointer => |info| switch (info.size) { - .One => @compileError("TODO auto hash for single item pointers"), - .Many => @compileError("TODO auto hash for many item pointers"), - .C => @compileError("TODO auto hash C pointers"), - .Slice => { - const interval = std.math.max(1, key.len / 256); - var i: usize = 0; - var h = comptime rng.scalar(HashInt); - while (i < key.len) : (i += interval) { - h ^= autoHash(key[i], rng, HashInt); - } - return h; - }, - }, - - .Optional => @compileError("TODO auto hash for optionals"), - .Array => @compileError("TODO auto hash for arrays"), - .Vector => @compileError("TODO auto hash for vectors"), - .Struct => @compileError("TODO auto hash for structs"), - .Union => @compileError("TODO auto hash for unions"), - .ErrorUnion => @compileError("TODO auto hash for unions"), - } -} - -pub fn autoEql(a: var, b: @typeOf(a)) bool { - switch (@typeInfo(@typeOf(a))) { - .NoReturn, - .Opaque, - .Undefined, - .ArgTuple, - => @compileError("cannot test equality of this type"), - .Void, - .Null, - => return true, - .Bool, - .Int, - .Float, - .ComptimeFloat, - .ComptimeInt, - .EnumLiteral, - .Promise, - .Enum, - .BoundFn, - .Fn, - .ErrorSet, - .Type, - => return a == b, - - .Pointer => |info| switch (info.size) { - .One => @compileError("TODO auto eql for single item pointers"), - .Many => @compileError("TODO auto eql for many item pointers"), - .C => @compileError("TODO auto eql for C pointers"), - .Slice => { - if (a.len != b.len) return false; - for (a) |a_item, i| { - if (!autoEql(a_item, b[i])) return false; - } - return true; - }, - }, - - .Optional => @compileError("TODO auto eql for optionals"), - .Array => @compileError("TODO auto eql for arrays"), - .Struct => @compileError("TODO auto eql for structs"), - .Union => @compileError("TODO auto eql for unions"), - .ErrorUnion => @compileError("TODO auto eql for unions"), - .Vector => @compileError("TODO auto eql for vectors"), - } -} diff --git a/std/http/headers.zig b/std/http/headers.zig index 69ed494f3a..c588f2d055 100644 --- a/std/http/headers.zig +++ b/std/http/headers.zig @@ -102,9 +102,19 @@ test "HeaderEntry" { testing.expectEqualSlices(u8, "x", e.value); } +fn stringEql(a: []const u8, b: []const u8) bool { + if (a.len != b.len) return false; + if (a.ptr == b.ptr) return true; + return mem.compare(u8, a, b) == .Equal; +} + +fn stringHash(s: []const u8) u32 { + return @truncate(u32, std.hash.Wyhash.hash(0, s)); +} + const HeaderList = std.ArrayList(HeaderEntry); const HeaderIndexList = std.ArrayList(usize); -const HeaderIndex = std.AutoHashMap([]const u8, HeaderIndexList); +const HeaderIndex = std.HashMap([]const u8, HeaderIndexList, stringHash, stringEql); pub const Headers = struct { // the owned header field name is stored in the index as part of the key diff --git a/std/os.zig b/std/os.zig index 9ff2e8f87f..c2010bf6a9 100644 --- a/std/os.zig +++ b/std/os.zig @@ -133,6 +133,11 @@ fn getRandomBytesDevURandom(buf: []u8) !void { const fd = try openC(c"/dev/urandom", O_RDONLY | O_CLOEXEC, 0); defer close(fd); + const st = try fstat(fd); + if (!S_ISCHR(st.mode)) { + return error.NoDevice; + } + const stream = &std.fs.File.openHandle(fd).inStream().stream; stream.readNoEof(buf) catch return error.Unexpected; } diff --git a/std/os/bits/darwin.zig b/std/os/bits/darwin.zig index b8d229dbe9..483d4cda90 100644 --- a/std/os/bits/darwin.zig +++ b/std/os/bits/darwin.zig @@ -1116,3 +1116,62 @@ pub const stack_t = extern struct { ss_size: isize, ss_flags: i32, }; + +pub const S_IFMT = 0o170000; + +pub const S_IFIFO = 0o010000; +pub const S_IFCHR = 0o020000; +pub const S_IFDIR = 0o040000; +pub const S_IFBLK = 0o060000; +pub const S_IFREG = 0o100000; +pub const S_IFLNK = 0o120000; +pub const S_IFSOCK = 0o140000; +pub const S_IFWHT = 0o160000; + +pub const S_ISUID = 0o4000; +pub const S_ISGID = 0o2000; +pub const S_ISVTX = 0o1000; +pub const S_IRWXU = 0o700; +pub const S_IRUSR = 0o400; +pub const S_IWUSR = 0o200; +pub const S_IXUSR = 0o100; +pub const S_IRWXG = 0o070; +pub const S_IRGRP = 0o040; +pub const S_IWGRP = 0o020; +pub const S_IXGRP = 0o010; +pub const S_IRWXO = 0o007; +pub const S_IROTH = 0o004; +pub const S_IWOTH = 0o002; +pub const S_IXOTH = 0o001; + +pub fn S_ISFIFO(m: u32) bool { + return m & S_IFMT == S_IFIFO; +} + +pub fn S_ISCHR(m: u32) bool { + return m & S_IFMT == S_IFCHR; +} + +pub fn S_ISDIR(m: u32) bool { + return m & S_IFMT == S_IFDIR; +} + +pub fn S_ISBLK(m: u32) bool { + return m & S_IFMT == S_IFBLK; +} + +pub fn S_ISREG(m: u32) bool { + return m & S_IFMT == S_IFREG; +} + +pub fn S_ISLNK(m: u32) bool { + return m & S_IFMT == S_IFLNK; +} + +pub fn S_ISSOCK(m: u32) bool { + return m & S_IFMT == S_IFSOCK; +} + +pub fn S_IWHT(m: u32) bool { + return m & S_IFMT == S_IFWHT; +} diff --git a/std/os/bits/freebsd.zig b/std/os/bits/freebsd.zig index 198857983e..45432a6c07 100644 --- a/std/os/bits/freebsd.zig +++ b/std/os/bits/freebsd.zig @@ -876,3 +876,62 @@ pub const stack_t = extern struct { ss_size: isize, ss_flags: i32, }; + +pub const S_IFMT = 0o170000; + +pub const S_IFIFO = 0o010000; +pub const S_IFCHR = 0o020000; +pub const S_IFDIR = 0o040000; +pub const S_IFBLK = 0o060000; +pub const S_IFREG = 0o100000; +pub const S_IFLNK = 0o120000; +pub const S_IFSOCK = 0o140000; +pub const S_IFWHT = 0o160000; + +pub const S_ISUID = 0o4000; +pub const S_ISGID = 0o2000; +pub const S_ISVTX = 0o1000; +pub const S_IRWXU = 0o700; +pub const S_IRUSR = 0o400; +pub const S_IWUSR = 0o200; +pub const S_IXUSR = 0o100; +pub const S_IRWXG = 0o070; +pub const S_IRGRP = 0o040; +pub const S_IWGRP = 0o020; +pub const S_IXGRP = 0o010; +pub const S_IRWXO = 0o007; +pub const S_IROTH = 0o004; +pub const S_IWOTH = 0o002; +pub const S_IXOTH = 0o001; + +pub fn S_ISFIFO(m: u32) bool { + return m & S_IFMT == S_IFIFO; +} + +pub fn S_ISCHR(m: u32) bool { + return m & S_IFMT == S_IFCHR; +} + +pub fn S_ISDIR(m: u32) bool { + return m & S_IFMT == S_IFDIR; +} + +pub fn S_ISBLK(m: u32) bool { + return m & S_IFMT == S_IFBLK; +} + +pub fn S_ISREG(m: u32) bool { + return m & S_IFMT == S_IFREG; +} + +pub fn S_ISLNK(m: u32) bool { + return m & S_IFMT == S_IFLNK; +} + +pub fn S_ISSOCK(m: u32) bool { + return m & S_IFMT == S_IFSOCK; +} + +pub fn S_IWHT(m: u32) bool { + return m & S_IFMT == S_IFWHT; +} diff --git a/std/os/darwin.zig b/std/os/darwin.zig index 67ce9a06cf..c2b6801e22 100644 --- a/std/os/darwin.zig +++ b/std/os/darwin.zig @@ -5,3 +5,4 @@ pub const is_the_target = switch (builtin.os) { else => false, }; pub usingnamespace std.c; +pub usingnamespace @import("bits.zig"); \ No newline at end of file diff --git a/std/os/freebsd.zig b/std/os/freebsd.zig index d418ccd415..e9efe64920 100644 --- a/std/os/freebsd.zig +++ b/std/os/freebsd.zig @@ -2,3 +2,4 @@ const std = @import("../std.zig"); const builtin = @import("builtin"); pub const is_the_target = builtin.os == .freebsd; pub usingnamespace std.c; +pub usingnamespace @import("bits.zig"); \ No newline at end of file diff --git a/std/rb.zig b/std/rb.zig index b5935a2eac..0b84950544 100644 --- a/std/rb.zig +++ b/std/rb.zig @@ -234,10 +234,13 @@ pub const Tree = struct { return null; } + /// lookup searches for the value of key, using binary search. It will + /// return a pointer to the node if it is there, otherwise it will return null. + /// Complexity guaranteed O(log n), where n is the number of nodes book-kept + /// by tree. pub fn lookup(tree: *Tree, key: *Node) ?*Node { - var parent: *Node = undefined; + var parent: ?*Node = undefined; var is_left: bool = undefined; - return doLookup(key, tree, &parent, &is_left); } @@ -545,3 +548,47 @@ test "rb" { num = testGetNumber(num.node.next().?); } } + + +test "inserting and looking up" { + var tree: Tree = undefined; + tree.init(testCompare); + var number: testNumber = undefined; + number.value = 1000; + _ = tree.insert(&number.node); + var dup: testNumber = undefined; + //Assert that tuples with identical value fields finds the same pointer + dup.value = 1000; + assert(tree.lookup(&dup.node) == &number.node); + //Assert that tuples with identical values do not clobber when inserted. + _ = tree.insert(&dup.node); + assert(tree.lookup(&dup.node) == &number.node); + assert(tree.lookup(&number.node) != &dup.node); + assert(testGetNumber(tree.lookup(&dup.node).?).value == testGetNumber(&dup.node).value); + //Assert that if looking for a non-existing value, return null. + var non_existing_value: testNumber = undefined; + non_existing_value.value = 1234; + assert(tree.lookup(&non_existing_value.node) == null); +} + +test "multiple inserts, followed by calling first and last" { + var tree: Tree = undefined; + tree.init(testCompare); + var zeroth: testNumber = undefined; + zeroth.value = 0; + var first: testNumber = undefined; + first.value = 1; + var second: testNumber = undefined; + second.value = 2; + var third: testNumber = undefined; + third.value = 3; + _ = tree.insert(&zeroth.node); + _ = tree.insert(&first.node); + _ = tree.insert(&second.node); + _ = tree.insert(&third.node); + assert(testGetNumber(tree.first().?).value == 0); + assert(testGetNumber(tree.last().?).value == 3); + var lookupNode: testNumber = undefined; + lookupNode.value = 3; + assert(tree.lookup(&lookupNode.node) == &third.node); +}