switch from "id" to "nonce"

mainly this addresses the following use case:

1. Someone creates a template with build.zig.zon, id field included
   (note that zig init does not create this problem since it generates
   fresh id every time it runs).
2. User A uses the template, changing package name to "example" but not
   id field.
3. User B uses the same template, changing package name also to
   "example", also not changing the id field.

Here, both packages have unintentional conflicting logical ids.

By making the field a combination of name checksum + random id, this
accident is avoided. "nonce" is an OK name for this.

Also relaxes errors on remote packages when using `zig fetch`.
This commit is contained in:
Andrew Kelley 2025-02-25 17:26:19 -08:00
parent a70307e7ff
commit 0fc7c9f57c
7 changed files with 100 additions and 59 deletions

View file

@ -12,5 +12,5 @@
}, },
}, },
.paths = .{""}, .paths = .{""},
.id = 0x1cb6, .nonce = 0xc1ce10810000f013,
} }

View file

@ -22,21 +22,30 @@ Zig package namespace.
Must be a valid bare Zig identifier (don't `@` me), limited to 32 bytes. Must be a valid bare Zig identifier (don't `@` me), limited to 32 bytes.
### `id` ### `nonce`
Together with name, this represents a globally unique package identifier. This Together with name, this represents a globally unique package identifier. This
field should be initialized with a 16-bit random number when the package is field is auto-initialized by the toolchain when the package is first created,
first created, and then *never change*. This allows Zig to unambiguously detect and then *never changes*. This allows Zig to unambiguously detect when one
when one package is an updated version of another. package is an updated version of another.
When forking a Zig project, this id should be regenerated with a new random When forking a Zig project, this nonce should be regenerated if the upstream
number if the upstream project is still maintained. Otherwise, the fork is project is still maintained. Otherwise, the fork is *hostile*, attempting to
*hostile*, attempting to take control over the original project's identity. take control over the original project's identity. The nonce can be regenerated
by deleting the field and running `zig build`.
`0x0000` is invalid because it obviously means a random number wasn't used. This 64-bit integer is the combination of a 16-bit id component, a 32-bit
checksum, and 16 bits of reserved zeroes.
The id component within the nonce has these restrictions:
`0x0000` is reserved for legacy packages.
`0xffff` is reserved to represent "naked" packages. `0xffff` is reserved to represent "naked" packages.
The checksum is computed from `name` and serves to protect Zig users from
accidental id collisions.
### `version` ### `version`
String. Required. String. Required.

View file

@ -13,17 +13,18 @@
.version = "0.0.0", .version = "0.0.0",
// Together with name, this represents a globally unique package // Together with name, this represents a globally unique package
// identifier. This field should be initialized with a 16-bit random number // identifier. This field is generated by the Zig toolchain when the
// when the package is first created, and then *never change*. This allows // package is first created, and then *never changes*. This allows
// unambiguous detection when one package is an updated version of another. // unambiguous detection of one package being an updated version of
// another.
// //
// When forking a Zig project, this id should be regenerated with a new // When forking a Zig project, this id should be regenerated (delete the
// random number if the upstream project is still maintained. Otherwise, // field and run `zig build`) if the upstream project is still maintained.
// the fork is *hostile*, attempting to take control over the original // Otherwise, the fork is *hostile*, attempting to take control over the
// project's identity. Thus it is recommended to leave the comment on the // original project's identity. Thus it is recommended to leave the comment
// following line intact, so that it shows up in code reviews that modify // on the following line intact, so that it shows up in code reviews that
// the field. // modify the field.
.id = $i, // Changing this has security and trust implications. .nonce = $i, // Changing this has security and trust implications.
// Tracks the earliest Zig version that the package considers to be a // Tracks the earliest Zig version that the package considers to be a
// supported use case. // supported use case.

View file

@ -10,9 +10,29 @@ pub const multihash_len = 1 + 1 + Hash.Algo.digest_length;
pub const multihash_hex_digest_len = 2 * multihash_len; pub const multihash_hex_digest_len = 2 * multihash_len;
pub const MultiHashHexDigest = [multihash_hex_digest_len]u8; pub const MultiHashHexDigest = [multihash_hex_digest_len]u8;
pub fn randomId() u16 { pub const Nonce = packed struct(u64) {
return std.crypto.random.intRangeLessThan(u16, 0x0001, 0xffff); id: u16,
} reserved: u16 = 0,
checksum: u32,
pub fn generate(name: []const u8) Nonce {
return .{
.id = std.crypto.random.intRangeLessThan(u16, 0x0001, 0xffff),
.checksum = std.hash.Crc32.hash(name),
};
}
pub fn validate(n: Nonce, name: []const u8) bool {
switch (n.id) {
0x0000, 0xffff => return false,
else => return std.hash.Crc32.hash(name) == n.checksum,
}
}
pub fn int(n: Nonce) u64 {
return @bitCast(n);
}
};
/// A user-readable, file system safe hash that identifies an exact package /// A user-readable, file system safe hash that identifies an exact package
/// snapshot, including file contents. /// snapshot, including file contents.
@ -72,9 +92,10 @@ pub const Hash = struct {
} }
/// Produces "$name-$semver-$hashplus". /// Produces "$name-$semver-$hashplus".
/// * name is the name field from build.zig.zon, truncated at 32 bytes and must /// * name is the name field from build.zig.zon, asserted to be at most 32
/// be a valid zig identifier /// bytes and assumed be a valid zig identifier
/// * semver is the version field from build.zig.zon, truncated at 32 bytes /// * semver is the version field from build.zig.zon, asserted to be at
/// most 32 bytes
/// * hashplus is the following 39-byte array, base64 encoded using -_ to make /// * hashplus is the following 39-byte array, base64 encoded using -_ to make
/// it filesystem safe: /// it filesystem safe:
/// - (2 bytes) LE u16 Package ID /// - (2 bytes) LE u16 Package ID

View file

@ -44,6 +44,8 @@ omit_missing_hash_error: bool,
/// which specifies inclusion rules. This is intended to be true for the first /// which specifies inclusion rules. This is intended to be true for the first
/// fetch task and false for the recursive dependencies. /// fetch task and false for the recursive dependencies.
allow_missing_paths_field: bool, allow_missing_paths_field: bool,
allow_missing_nonce: bool,
allow_name_string: bool,
/// If true and URL points to a Git repository, will use the latest commit. /// If true and URL points to a Git repository, will use the latest commit.
use_latest_commit: bool, use_latest_commit: bool,
@ -372,7 +374,7 @@ pub fn run(f: *Fetch) RunError!void {
}; };
if (remote.hash) |expected_hash| { if (remote.hash) |expected_hash| {
var prefixed_pkg_sub_path_buffer: [100]u8 = undefined; var prefixed_pkg_sub_path_buffer: [Package.Hash.max_len + 2]u8 = undefined;
prefixed_pkg_sub_path_buffer[0] = 'p'; prefixed_pkg_sub_path_buffer[0] = 'p';
prefixed_pkg_sub_path_buffer[1] = fs.path.sep; prefixed_pkg_sub_path_buffer[1] = fs.path.sep;
const hash_slice = expected_hash.toSlice(); const hash_slice = expected_hash.toSlice();
@ -647,8 +649,8 @@ fn loadManifest(f: *Fetch, pkg_root: Cache.Path) RunError!void {
f.manifest = try Manifest.parse(arena, ast.*, .{ f.manifest = try Manifest.parse(arena, ast.*, .{
.allow_missing_paths_field = f.allow_missing_paths_field, .allow_missing_paths_field = f.allow_missing_paths_field,
.allow_missing_id = f.allow_missing_paths_field, .allow_missing_nonce = f.allow_missing_nonce,
.allow_name_string = f.allow_missing_paths_field, .allow_name_string = f.allow_name_string,
}); });
const manifest = &f.manifest.?; const manifest = &f.manifest.?;
@ -750,6 +752,8 @@ fn queueJobsForDeps(f: *Fetch) RunError!void {
.job_queue = f.job_queue, .job_queue = f.job_queue,
.omit_missing_hash_error = false, .omit_missing_hash_error = false,
.allow_missing_paths_field = true, .allow_missing_paths_field = true,
.allow_missing_nonce = true,
.allow_name_string = true,
.use_latest_commit = false, .use_latest_commit = false,
.package_root = undefined, .package_root = undefined,
@ -2319,6 +2323,8 @@ const TestFetchBuilder = struct {
.job_queue = &self.job_queue, .job_queue = &self.job_queue,
.omit_missing_hash_error = true, .omit_missing_hash_error = true,
.allow_missing_paths_field = false, .allow_missing_paths_field = false,
.allow_missing_nonce = true, // so we can keep using the old testdata .tar.gz
.allow_name_string = true, // so we can keep using the old testdata .tar.gz
.use_latest_commit = true, .use_latest_commit = true,
.package_root = undefined, .package_root = undefined,

View file

@ -52,7 +52,7 @@ pub const ParseOptions = struct {
/// Deprecated, to be removed after 0.14.0 is tagged. /// Deprecated, to be removed after 0.14.0 is tagged.
allow_name_string: bool = true, allow_name_string: bool = true,
/// Deprecated, to be removed after 0.14.0 is tagged. /// Deprecated, to be removed after 0.14.0 is tagged.
allow_missing_id: bool = true, allow_missing_nonce: bool = true,
}; };
pub const Error = Allocator.Error; pub const Error = Allocator.Error;
@ -81,7 +81,7 @@ pub fn parse(gpa: Allocator, ast: Ast, options: ParseOptions) Error!Manifest {
.paths = .{}, .paths = .{},
.allow_missing_paths_field = options.allow_missing_paths_field, .allow_missing_paths_field = options.allow_missing_paths_field,
.allow_name_string = options.allow_name_string, .allow_name_string = options.allow_name_string,
.allow_missing_id = options.allow_missing_id, .allow_missing_nonce = options.allow_missing_nonce,
.minimum_zig_version = null, .minimum_zig_version = null,
.buf = .{}, .buf = .{},
}; };
@ -157,7 +157,7 @@ const Parse = struct {
paths: std.StringArrayHashMapUnmanaged(void), paths: std.StringArrayHashMapUnmanaged(void),
allow_missing_paths_field: bool, allow_missing_paths_field: bool,
allow_name_string: bool, allow_name_string: bool,
allow_missing_id: bool, allow_missing_nonce: bool,
minimum_zig_version: ?std.SemanticVersion, minimum_zig_version: ?std.SemanticVersion,
const InnerError = error{ ParseFailure, OutOfMemory }; const InnerError = error{ ParseFailure, OutOfMemory };
@ -175,7 +175,7 @@ const Parse = struct {
var have_name = false; var have_name = false;
var have_version = false; var have_version = false;
var have_included_paths = false; var have_included_paths = false;
var have_id = false; var nonce: ?Package.Nonce = null;
for (struct_init.ast.fields) |field_init| { for (struct_init.ast.fields) |field_init| {
const name_token = ast.firstToken(field_init) - 2; const name_token = ast.firstToken(field_init) - 2;
@ -192,9 +192,8 @@ const Parse = struct {
} else if (mem.eql(u8, field_name, "name")) { } else if (mem.eql(u8, field_name, "name")) {
p.name = try parseName(p, field_init); p.name = try parseName(p, field_init);
have_name = true; have_name = true;
} else if (mem.eql(u8, field_name, "id")) { } else if (mem.eql(u8, field_name, "nonce")) {
p.id = try parseId(p, field_init); nonce = try parseNonce(p, field_init);
have_id = true;
} else if (mem.eql(u8, field_name, "version")) { } else if (mem.eql(u8, field_name, "version")) {
p.version_node = field_init; p.version_node = field_init;
const version_text = try parseString(p, field_init); const version_text = try parseString(p, field_init);
@ -218,14 +217,23 @@ const Parse = struct {
} }
} }
if (!have_id and !p.allow_missing_id) {
try appendError(p, main_token, "missing top-level 'id' field; suggested value: 0x{x}", .{
Package.randomId(),
});
}
if (!have_name) { if (!have_name) {
try appendError(p, main_token, "missing top-level 'name' field", .{}); try appendError(p, main_token, "missing top-level 'name' field", .{});
} else {
if (nonce) |n| {
if (!n.validate(p.name)) {
return fail(p, main_token, "invalid nonce: 0x{x}; if this is a new or forked package, use this value: 0x{x}", .{
n.int(), Package.Nonce.generate(p.name).int(),
});
}
p.id = n.id;
} else if (!p.allow_missing_nonce) {
try appendError(p, main_token, "missing top-level 'nonce' field; suggested value: 0x{x}", .{
Package.Nonce.generate(p.name).int(),
});
} else {
p.id = 0;
}
} }
if (!have_version) { if (!have_version) {
@ -377,7 +385,7 @@ const Parse = struct {
} }
} }
fn parseId(p: *Parse, node: Ast.Node.Index) !u16 { fn parseNonce(p: *Parse, node: Ast.Node.Index) !Package.Nonce {
const ast = p.ast; const ast = p.ast;
const node_tags = ast.nodes.items(.tag); const node_tags = ast.nodes.items(.tag);
const main_tokens = ast.nodes.items(.main_token); const main_tokens = ast.nodes.items(.main_token);
@ -387,20 +395,12 @@ const Parse = struct {
} }
const token_bytes = ast.tokenSlice(main_token); const token_bytes = ast.tokenSlice(main_token);
const parsed = std.zig.parseNumberLiteral(token_bytes); const parsed = std.zig.parseNumberLiteral(token_bytes);
const n = switch (parsed) { switch (parsed) {
.int => |n| n, .int => |n| return @bitCast(n),
.big_int, .float => return fail(p, main_token, "expected u16 integer literal, found {s}", .{ .big_int, .float => return fail(p, main_token, "expected u64 integer literal, found {s}", .{
@tagName(parsed), @tagName(parsed),
}), }),
.failure => |err| return fail(p, main_token, "bad integer literal: {s}", .{@tagName(err)}), .failure => |err| return fail(p, main_token, "bad integer literal: {s}", .{@tagName(err)}),
};
const casted = std.math.cast(u16, n) orelse
return fail(p, main_token, "integer value {d} does not fit into u16", .{n});
switch (casted) {
0x0000, 0xffff => return fail(p, main_token, "id value 0x{x} reserved; use 0x{x} instead", .{
casted, Package.randomId(),
}),
else => return casted,
} }
} }

View file

@ -4752,10 +4752,10 @@ fn cmdInit(gpa: Allocator, arena: Allocator, args: []const []const u8) !void {
}; };
var ok_count: usize = 0; var ok_count: usize = 0;
const id = Package.randomId(); const nonce: Package.Nonce = .generate(sanitized_root_name);
for (template_paths) |template_path| { for (template_paths) |template_path| {
if (templates.write(arena, fs.cwd(), sanitized_root_name, template_path, id)) |_| { if (templates.write(arena, fs.cwd(), sanitized_root_name, template_path, nonce)) |_| {
std.log.info("created {s}", .{template_path}); std.log.info("created {s}", .{template_path});
ok_count += 1; ok_count += 1;
} else |err| switch (err) { } else |err| switch (err) {
@ -5225,6 +5225,8 @@ fn cmdBuild(gpa: Allocator, arena: Allocator, args: []const []const u8) !void {
.job_queue = &job_queue, .job_queue = &job_queue,
.omit_missing_hash_error = true, .omit_missing_hash_error = true,
.allow_missing_paths_field = false, .allow_missing_paths_field = false,
.allow_missing_nonce = false,
.allow_name_string = false,
.use_latest_commit = false, .use_latest_commit = false,
.package_root = undefined, .package_root = undefined,
@ -7125,6 +7127,8 @@ fn cmdFetch(
.job_queue = &job_queue, .job_queue = &job_queue,
.omit_missing_hash_error = true, .omit_missing_hash_error = true,
.allow_missing_paths_field = false, .allow_missing_paths_field = false,
.allow_missing_nonce = true,
.allow_name_string = true,
.use_latest_commit = true, .use_latest_commit = true,
.package_root = undefined, .package_root = undefined,
@ -7464,10 +7468,10 @@ fn loadManifest(
0, 0,
) catch |err| switch (err) { ) catch |err| switch (err) {
error.FileNotFound => { error.FileNotFound => {
const id = Package.randomId(); const nonce: Package.Nonce = .generate(options.root_name);
var templates = findTemplates(gpa, arena); var templates = findTemplates(gpa, arena);
defer templates.deinit(); defer templates.deinit();
templates.write(arena, options.dir, options.root_name, Package.Manifest.basename, id) catch |e| { templates.write(arena, options.dir, options.root_name, Package.Manifest.basename, nonce) catch |e| {
fatal("unable to write {s}: {s}", .{ fatal("unable to write {s}: {s}", .{
Package.Manifest.basename, @errorName(e), Package.Manifest.basename, @errorName(e),
}); });
@ -7525,7 +7529,7 @@ const Templates = struct {
out_dir: fs.Dir, out_dir: fs.Dir,
root_name: []const u8, root_name: []const u8,
template_path: []const u8, template_path: []const u8,
id: u16, nonce: Package.Nonce,
) !void { ) !void {
if (fs.path.dirname(template_path)) |dirname| { if (fs.path.dirname(template_path)) |dirname| {
out_dir.makePath(dirname) catch |err| { out_dir.makePath(dirname) catch |err| {
@ -7551,7 +7555,7 @@ const Templates = struct {
state = .start; state = .start;
}, },
'i' => { 'i' => {
try templates.buffer.writer().print("0x{x}", .{id}); try templates.buffer.writer().print("0x{x}", .{nonce.int()});
state = .start; state = .start;
}, },
'v' => { 'v' => {