mirror of
https://github.com/rust-lang/rust.git
synced 2024-11-27 01:04:03 +00:00
fb5002d68a
`DefId` uses different field orders on 64-bit big-endian vs. others, in order to optimize its `Hash` implementation. However, that also made it derive different lexical ordering for `PartialOrd` and `Ord`. That caused spurious differences wherever `DefId`s are sorted, like the candidate sources list in `report_method_error`. Now we manually implement `PartialOrd` and `Ord` on 64-bit big-endian to match the same lexical ordering as other targets, fixing at least one test, `src/test/ui/methods/method-ambig-two-traits-cross-crate.rs`.
474 lines
16 KiB
Rust
474 lines
16 KiB
Rust
use crate::HashStableContext;
|
|
use rustc_data_structures::fingerprint::Fingerprint;
|
|
use rustc_data_structures::stable_hasher::{HashStable, StableHasher, ToStableHashKey};
|
|
use rustc_data_structures::AtomicRef;
|
|
use rustc_index::vec::Idx;
|
|
use rustc_macros::HashStable_Generic;
|
|
use rustc_serialize::{Decodable, Decoder, Encodable, Encoder};
|
|
use std::borrow::Borrow;
|
|
use std::fmt;
|
|
use std::hash::{Hash, Hasher};
|
|
|
|
rustc_index::newtype_index! {
|
|
pub struct CrateNum {
|
|
ENCODABLE = custom
|
|
DEBUG_FORMAT = "crate{}"
|
|
}
|
|
}
|
|
|
|
/// Item definitions in the currently-compiled crate would have the `CrateNum`
|
|
/// `LOCAL_CRATE` in their `DefId`.
|
|
pub const LOCAL_CRATE: CrateNum = CrateNum::from_u32(0);
|
|
|
|
impl CrateNum {
|
|
#[inline]
|
|
pub fn new(x: usize) -> CrateNum {
|
|
CrateNum::from_usize(x)
|
|
}
|
|
|
|
#[inline]
|
|
pub fn as_def_id(self) -> DefId {
|
|
DefId { krate: self, index: CRATE_DEF_INDEX }
|
|
}
|
|
}
|
|
|
|
impl fmt::Display for CrateNum {
|
|
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
|
|
fmt::Display::fmt(&self.private, f)
|
|
}
|
|
}
|
|
|
|
/// As a local identifier, a `CrateNum` is only meaningful within its context, e.g. within a tcx.
|
|
/// Therefore, make sure to include the context when encode a `CrateNum`.
|
|
impl<E: Encoder> Encodable<E> for CrateNum {
|
|
default fn encode(&self, s: &mut E) {
|
|
s.emit_u32(self.as_u32());
|
|
}
|
|
}
|
|
|
|
impl<D: Decoder> Decodable<D> for CrateNum {
|
|
default fn decode(d: &mut D) -> CrateNum {
|
|
CrateNum::from_u32(d.read_u32())
|
|
}
|
|
}
|
|
|
|
/// A `DefPathHash` is a fixed-size representation of a `DefPath` that is
|
|
/// stable across crate and compilation session boundaries. It consists of two
|
|
/// separate 64-bit hashes. The first uniquely identifies the crate this
|
|
/// `DefPathHash` originates from (see [StableCrateId]), and the second
|
|
/// uniquely identifies the corresponding `DefPath` within that crate. Together
|
|
/// they form a unique identifier within an entire crate graph.
|
|
///
|
|
/// There is a very small chance of hash collisions, which would mean that two
|
|
/// different `DefPath`s map to the same `DefPathHash`. Proceeding compilation
|
|
/// with such a hash collision would very probably lead to an ICE, and in the
|
|
/// worst case lead to a silent mis-compilation. The compiler therefore actively
|
|
/// and exhaustively checks for such hash collisions and aborts compilation if
|
|
/// it finds one.
|
|
///
|
|
/// `DefPathHash` uses 64-bit hashes for both the crate-id part and the
|
|
/// crate-internal part, even though it is likely that there are many more
|
|
/// `LocalDefId`s in a single crate than there are individual crates in a crate
|
|
/// graph. Since we use the same number of bits in both cases, the collision
|
|
/// probability for the crate-local part will be quite a bit higher (though
|
|
/// still very small).
|
|
///
|
|
/// This imbalance is not by accident: A hash collision in the
|
|
/// crate-local part of a `DefPathHash` will be detected and reported while
|
|
/// compiling the crate in question. Such a collision does not depend on
|
|
/// outside factors and can be easily fixed by the crate maintainer (e.g. by
|
|
/// renaming the item in question or by bumping the crate version in a harmless
|
|
/// way).
|
|
///
|
|
/// A collision between crate-id hashes on the other hand is harder to fix
|
|
/// because it depends on the set of crates in the entire crate graph of a
|
|
/// compilation session. Again, using the same crate with a different version
|
|
/// number would fix the issue with a high probability -- but that might be
|
|
/// easier said then done if the crates in questions are dependencies of
|
|
/// third-party crates.
|
|
///
|
|
/// That being said, given a high quality hash function, the collision
|
|
/// probabilities in question are very small. For example, for a big crate like
|
|
/// `rustc_middle` (with ~50000 `LocalDefId`s as of the time of writing) there
|
|
/// is a probability of roughly 1 in 14,750,000,000 of a crate-internal
|
|
/// collision occurring. For a big crate graph with 1000 crates in it, there is
|
|
/// a probability of 1 in 36,890,000,000,000 of a `StableCrateId` collision.
|
|
#[derive(Copy, Clone, Hash, PartialEq, Eq, PartialOrd, Ord, Debug)]
|
|
#[derive(HashStable_Generic, Encodable, Decodable)]
|
|
pub struct DefPathHash(pub Fingerprint);
|
|
|
|
impl DefPathHash {
|
|
/// Returns the [StableCrateId] identifying the crate this [DefPathHash]
|
|
/// originates from.
|
|
#[inline]
|
|
pub fn stable_crate_id(&self) -> StableCrateId {
|
|
StableCrateId(self.0.as_value().0)
|
|
}
|
|
|
|
/// Returns the crate-local part of the [DefPathHash].
|
|
///
|
|
/// Used for tests.
|
|
#[inline]
|
|
pub fn local_hash(&self) -> u64 {
|
|
self.0.as_value().1
|
|
}
|
|
|
|
/// Builds a new [DefPathHash] with the given [StableCrateId] and
|
|
/// `local_hash`, where `local_hash` must be unique within its crate.
|
|
pub fn new(stable_crate_id: StableCrateId, local_hash: u64) -> DefPathHash {
|
|
DefPathHash(Fingerprint::new(stable_crate_id.0, local_hash))
|
|
}
|
|
}
|
|
|
|
impl Borrow<Fingerprint> for DefPathHash {
|
|
#[inline]
|
|
fn borrow(&self) -> &Fingerprint {
|
|
&self.0
|
|
}
|
|
}
|
|
|
|
/// A [`StableCrateId`] is a 64-bit hash of a crate name, together with all
|
|
/// `-Cmetadata` arguments, and some other data. It is to [`CrateNum`] what [`DefPathHash`] is to
|
|
/// [`DefId`]. It is stable across compilation sessions.
|
|
///
|
|
/// Since the ID is a hash value, there is a small chance that two crates
|
|
/// end up with the same [`StableCrateId`]. The compiler will check for such
|
|
/// collisions when loading crates and abort compilation in order to avoid
|
|
/// further trouble.
|
|
///
|
|
/// For more information on the possibility of hash collisions in rustc,
|
|
/// see the discussion in [`DefId`].
|
|
#[derive(Copy, Clone, Hash, PartialEq, Eq, PartialOrd, Ord, Debug)]
|
|
#[derive(HashStable_Generic, Encodable, Decodable)]
|
|
pub struct StableCrateId(pub(crate) u64);
|
|
|
|
impl StableCrateId {
|
|
pub fn to_u64(self) -> u64 {
|
|
self.0
|
|
}
|
|
|
|
/// Computes the stable ID for a crate with the given name and
|
|
/// `-Cmetadata` arguments.
|
|
pub fn new(crate_name: &str, is_exe: bool, mut metadata: Vec<String>) -> StableCrateId {
|
|
let mut hasher = StableHasher::new();
|
|
crate_name.hash(&mut hasher);
|
|
|
|
// We don't want the stable crate ID to depend on the order of
|
|
// -C metadata arguments, so sort them:
|
|
metadata.sort();
|
|
// Every distinct -C metadata value is only incorporated once:
|
|
metadata.dedup();
|
|
|
|
hasher.write(b"metadata");
|
|
for s in &metadata {
|
|
// Also incorporate the length of a metadata string, so that we generate
|
|
// different values for `-Cmetadata=ab -Cmetadata=c` and
|
|
// `-Cmetadata=a -Cmetadata=bc`
|
|
hasher.write_usize(s.len());
|
|
hasher.write(s.as_bytes());
|
|
}
|
|
|
|
// Also incorporate crate type, so that we don't get symbol conflicts when
|
|
// linking against a library of the same name, if this is an executable.
|
|
hasher.write(if is_exe { b"exe" } else { b"lib" });
|
|
|
|
// Also incorporate the rustc version. Otherwise, with -Zsymbol-mangling-version=v0
|
|
// and no -Cmetadata, symbols from the same crate compiled with different versions of
|
|
// rustc are named the same.
|
|
//
|
|
// RUSTC_FORCE_RUSTC_VERSION is used to inject rustc version information
|
|
// during testing.
|
|
if let Some(val) = std::env::var_os("RUSTC_FORCE_RUSTC_VERSION") {
|
|
hasher.write(val.to_string_lossy().into_owned().as_bytes())
|
|
} else {
|
|
hasher.write(option_env!("CFG_VERSION").unwrap_or("unknown version").as_bytes());
|
|
}
|
|
|
|
StableCrateId(hasher.finish())
|
|
}
|
|
}
|
|
|
|
rustc_index::newtype_index! {
|
|
/// A DefIndex is an index into the hir-map for a crate, identifying a
|
|
/// particular definition. It should really be considered an interned
|
|
/// shorthand for a particular DefPath.
|
|
pub struct DefIndex {
|
|
ENCODABLE = custom // (only encodable in metadata)
|
|
|
|
DEBUG_FORMAT = "DefIndex({})",
|
|
/// The crate root is always assigned index 0 by the AST Map code,
|
|
/// thanks to `NodeCollector::new`.
|
|
const CRATE_DEF_INDEX = 0,
|
|
}
|
|
}
|
|
|
|
impl<E: Encoder> Encodable<E> for DefIndex {
|
|
default fn encode(&self, _: &mut E) {
|
|
panic!("cannot encode `DefIndex` with `{}`", std::any::type_name::<E>());
|
|
}
|
|
}
|
|
|
|
impl<D: Decoder> Decodable<D> for DefIndex {
|
|
default fn decode(_: &mut D) -> DefIndex {
|
|
panic!("cannot decode `DefIndex` with `{}`", std::any::type_name::<D>());
|
|
}
|
|
}
|
|
|
|
/// A `DefId` identifies a particular *definition*, by combining a crate
|
|
/// index and a def index.
|
|
///
|
|
/// You can create a `DefId` from a `LocalDefId` using `local_def_id.to_def_id()`.
|
|
#[derive(Clone, PartialEq, Eq, Copy)]
|
|
// Don't derive order on 64-bit big-endian, so we can be consistent regardless of field order.
|
|
#[cfg_attr(not(all(target_pointer_width = "64", target_endian = "big")), derive(PartialOrd, Ord))]
|
|
// On below-64 bit systems we can simply use the derived `Hash` impl
|
|
#[cfg_attr(not(target_pointer_width = "64"), derive(Hash))]
|
|
#[repr(C)]
|
|
#[rustc_pass_by_value]
|
|
// We guarantee field order. Note that the order is essential here, see below why.
|
|
pub struct DefId {
|
|
// cfg-ing the order of fields so that the `DefIndex` which is high entropy always ends up in
|
|
// the lower bits no matter the endianness. This allows the compiler to turn that `Hash` impl
|
|
// into a direct call to 'u64::hash(_)`.
|
|
#[cfg(not(all(target_pointer_width = "64", target_endian = "big")))]
|
|
pub index: DefIndex,
|
|
pub krate: CrateNum,
|
|
#[cfg(all(target_pointer_width = "64", target_endian = "big"))]
|
|
pub index: DefIndex,
|
|
}
|
|
|
|
// On 64-bit systems, we can hash the whole `DefId` as one `u64` instead of two `u32`s. This
|
|
// improves performance without impairing `FxHash` quality. So the below code gets compiled to a
|
|
// noop on little endian systems because the memory layout of `DefId` is as follows:
|
|
//
|
|
// ```
|
|
// +-1--------------31-+-32-------------63-+
|
|
// ! index ! krate !
|
|
// +-------------------+-------------------+
|
|
// ```
|
|
//
|
|
// The order here has direct impact on `FxHash` quality because we have far more `DefIndex` per
|
|
// crate than we have `Crate`s within one compilation. Or in other words, this arrangement puts
|
|
// more entropy in the low bits than the high bits. The reason this matters is that `FxHash`, which
|
|
// is used throughout rustc, has problems distributing the entropy from the high bits, so reversing
|
|
// the order would lead to a large number of collisions and thus far worse performance.
|
|
//
|
|
// On 64-bit big-endian systems, this compiles to a 64-bit rotation by 32 bits, which is still
|
|
// faster than another `FxHash` round.
|
|
#[cfg(target_pointer_width = "64")]
|
|
impl Hash for DefId {
|
|
fn hash<H: Hasher>(&self, h: &mut H) {
|
|
(((self.krate.as_u32() as u64) << 32) | (self.index.as_u32() as u64)).hash(h)
|
|
}
|
|
}
|
|
|
|
// Implement the same comparison as derived with the other field order.
|
|
#[cfg(all(target_pointer_width = "64", target_endian = "big"))]
|
|
impl Ord for DefId {
|
|
#[inline]
|
|
fn cmp(&self, other: &DefId) -> std::cmp::Ordering {
|
|
Ord::cmp(&(self.index, self.krate), &(other.index, other.krate))
|
|
}
|
|
}
|
|
#[cfg(all(target_pointer_width = "64", target_endian = "big"))]
|
|
impl PartialOrd for DefId {
|
|
#[inline]
|
|
fn partial_cmp(&self, other: &DefId) -> Option<std::cmp::Ordering> {
|
|
Some(Ord::cmp(self, other))
|
|
}
|
|
}
|
|
|
|
impl DefId {
|
|
/// Makes a local `DefId` from the given `DefIndex`.
|
|
#[inline]
|
|
pub fn local(index: DefIndex) -> DefId {
|
|
DefId { krate: LOCAL_CRATE, index }
|
|
}
|
|
|
|
/// Returns whether the item is defined in the crate currently being compiled.
|
|
#[inline]
|
|
pub fn is_local(self) -> bool {
|
|
self.krate == LOCAL_CRATE
|
|
}
|
|
|
|
#[inline]
|
|
pub fn as_local(self) -> Option<LocalDefId> {
|
|
if self.is_local() { Some(LocalDefId { local_def_index: self.index }) } else { None }
|
|
}
|
|
|
|
#[inline]
|
|
#[track_caller]
|
|
pub fn expect_local(self) -> LocalDefId {
|
|
// NOTE: `match` below is required to apply `#[track_caller]`,
|
|
// i.e. don't use closures.
|
|
match self.as_local() {
|
|
Some(local_def_id) => local_def_id,
|
|
None => panic!("DefId::expect_local: `{:?}` isn't local", self),
|
|
}
|
|
}
|
|
|
|
#[inline]
|
|
pub fn is_crate_root(self) -> bool {
|
|
self.index == CRATE_DEF_INDEX
|
|
}
|
|
|
|
#[inline]
|
|
pub fn as_crate_root(self) -> Option<CrateNum> {
|
|
if self.is_crate_root() { Some(self.krate) } else { None }
|
|
}
|
|
|
|
#[inline]
|
|
pub fn is_top_level_module(self) -> bool {
|
|
self.is_local() && self.is_crate_root()
|
|
}
|
|
}
|
|
|
|
impl From<LocalDefId> for DefId {
|
|
fn from(local: LocalDefId) -> DefId {
|
|
local.to_def_id()
|
|
}
|
|
}
|
|
|
|
impl<E: Encoder> Encodable<E> for DefId {
|
|
default fn encode(&self, s: &mut E) {
|
|
self.krate.encode(s);
|
|
self.index.encode(s);
|
|
}
|
|
}
|
|
|
|
impl<D: Decoder> Decodable<D> for DefId {
|
|
default fn decode(d: &mut D) -> DefId {
|
|
DefId { krate: Decodable::decode(d), index: Decodable::decode(d) }
|
|
}
|
|
}
|
|
|
|
pub fn default_def_id_debug(def_id: DefId, f: &mut fmt::Formatter<'_>) -> fmt::Result {
|
|
f.debug_struct("DefId").field("krate", &def_id.krate).field("index", &def_id.index).finish()
|
|
}
|
|
|
|
pub static DEF_ID_DEBUG: AtomicRef<fn(DefId, &mut fmt::Formatter<'_>) -> fmt::Result> =
|
|
AtomicRef::new(&(default_def_id_debug as fn(_, &mut fmt::Formatter<'_>) -> _));
|
|
|
|
impl fmt::Debug for DefId {
|
|
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
|
|
(*DEF_ID_DEBUG)(*self, f)
|
|
}
|
|
}
|
|
|
|
rustc_data_structures::define_id_collections!(DefIdMap, DefIdSet, DefIdMapEntry, DefId);
|
|
|
|
/// A `LocalDefId` is equivalent to a `DefId` with `krate == LOCAL_CRATE`. Since
|
|
/// we encode this information in the type, we can ensure at compile time that
|
|
/// no `DefId`s from upstream crates get thrown into the mix. There are quite a
|
|
/// few cases where we know that only `DefId`s from the local crate are expected;
|
|
/// a `DefId` from a different crate would signify a bug somewhere. This
|
|
/// is when `LocalDefId` comes in handy.
|
|
#[derive(Clone, Copy, PartialEq, Eq, Hash)]
|
|
pub struct LocalDefId {
|
|
pub local_def_index: DefIndex,
|
|
}
|
|
|
|
// To ensure correctness of incremental compilation,
|
|
// `LocalDefId` must not implement `Ord` or `PartialOrd`.
|
|
// See https://github.com/rust-lang/rust/issues/90317.
|
|
impl !Ord for LocalDefId {}
|
|
impl !PartialOrd for LocalDefId {}
|
|
|
|
pub const CRATE_DEF_ID: LocalDefId = LocalDefId { local_def_index: CRATE_DEF_INDEX };
|
|
|
|
impl Idx for LocalDefId {
|
|
#[inline]
|
|
fn new(idx: usize) -> Self {
|
|
LocalDefId { local_def_index: Idx::new(idx) }
|
|
}
|
|
#[inline]
|
|
fn index(self) -> usize {
|
|
self.local_def_index.index()
|
|
}
|
|
}
|
|
|
|
impl LocalDefId {
|
|
#[inline]
|
|
pub fn to_def_id(self) -> DefId {
|
|
DefId { krate: LOCAL_CRATE, index: self.local_def_index }
|
|
}
|
|
|
|
#[inline]
|
|
pub fn is_top_level_module(self) -> bool {
|
|
self == CRATE_DEF_ID
|
|
}
|
|
}
|
|
|
|
impl fmt::Debug for LocalDefId {
|
|
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
|
|
self.to_def_id().fmt(f)
|
|
}
|
|
}
|
|
|
|
impl<E: Encoder> Encodable<E> for LocalDefId {
|
|
fn encode(&self, s: &mut E) {
|
|
self.to_def_id().encode(s);
|
|
}
|
|
}
|
|
|
|
impl<D: Decoder> Decodable<D> for LocalDefId {
|
|
fn decode(d: &mut D) -> LocalDefId {
|
|
DefId::decode(d).expect_local()
|
|
}
|
|
}
|
|
|
|
rustc_data_structures::define_id_collections!(
|
|
LocalDefIdMap,
|
|
LocalDefIdSet,
|
|
LocalDefIdMapEntry,
|
|
LocalDefId
|
|
);
|
|
|
|
impl<CTX: HashStableContext> HashStable<CTX> for DefId {
|
|
#[inline]
|
|
fn hash_stable(&self, hcx: &mut CTX, hasher: &mut StableHasher) {
|
|
self.to_stable_hash_key(hcx).hash_stable(hcx, hasher);
|
|
}
|
|
}
|
|
|
|
impl<CTX: HashStableContext> HashStable<CTX> for LocalDefId {
|
|
#[inline]
|
|
fn hash_stable(&self, hcx: &mut CTX, hasher: &mut StableHasher) {
|
|
self.to_stable_hash_key(hcx).hash_stable(hcx, hasher);
|
|
}
|
|
}
|
|
|
|
impl<CTX: HashStableContext> HashStable<CTX> for CrateNum {
|
|
#[inline]
|
|
fn hash_stable(&self, hcx: &mut CTX, hasher: &mut StableHasher) {
|
|
self.to_stable_hash_key(hcx).hash_stable(hcx, hasher);
|
|
}
|
|
}
|
|
|
|
impl<CTX: HashStableContext> ToStableHashKey<CTX> for DefId {
|
|
type KeyType = DefPathHash;
|
|
|
|
#[inline]
|
|
fn to_stable_hash_key(&self, hcx: &CTX) -> DefPathHash {
|
|
hcx.def_path_hash(*self)
|
|
}
|
|
}
|
|
|
|
impl<CTX: HashStableContext> ToStableHashKey<CTX> for LocalDefId {
|
|
type KeyType = DefPathHash;
|
|
|
|
#[inline]
|
|
fn to_stable_hash_key(&self, hcx: &CTX) -> DefPathHash {
|
|
hcx.def_path_hash(self.to_def_id())
|
|
}
|
|
}
|
|
|
|
impl<CTX: HashStableContext> ToStableHashKey<CTX> for CrateNum {
|
|
type KeyType = DefPathHash;
|
|
|
|
#[inline]
|
|
fn to_stable_hash_key(&self, hcx: &CTX) -> DefPathHash {
|
|
self.as_def_id().to_stable_hash_key(hcx)
|
|
}
|
|
}
|