rust/src/librustc/ty/maps/job.rs

535 lines
18 KiB
Rust
Raw Normal View History

// Copyright 2017 The Rust Project Developers. See the COPYRIGHT
// file at the top-level directory of this distribution and at
// http://rust-lang.org/COPYRIGHT.
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
// except according to those terms.
#![allow(warnings)]
use std::mem;
use rustc_data_structures::sync::{Lock, LockGuard, Lrc, Weak};
use rustc_data_structures::OnDrop;
use syntax_pos::Span;
use ty::tls;
use ty::maps::Query;
use ty::maps::plumbing::CycleError;
use ty::context::TyCtxt;
use errors::Diagnostic;
use std::process;
use std::fmt;
use std::collections::HashSet;
#[cfg(parallel_queries)]
use {
rayon_core,
parking_lot::{Mutex, Condvar},
std::sync::atomic::Ordering,
std::thread,
std::iter,
std::iter::FromIterator,
syntax_pos::DUMMY_SP,
rustc_data_structures::stable_hasher::{StableHasherResult, StableHasher, HashStable},
};
2018-03-24 00:19:20 -05:00
/// Indicates the state of a query for a given key in a query map
pub(super) enum QueryResult<'tcx> {
2018-03-24 00:19:20 -05:00
/// An already executing query. The query job can be used to await for its completion
Started(Lrc<QueryJob<'tcx>>),
/// The query panicked. Queries trying to wait on this will raise a fatal error / silently panic
Poisoned,
}
/// A span and a query key
#[derive(Clone, Debug)]
2018-03-24 00:19:20 -05:00
pub struct QueryInfo<'tcx> {
2018-04-13 15:20:10 -05:00
/// The span for a reason this query was required
pub span: Span,
pub query: Query<'tcx>,
}
2018-03-24 00:19:20 -05:00
/// A object representing an active query job.
pub struct QueryJob<'tcx> {
2018-03-24 00:19:20 -05:00
pub info: QueryInfo<'tcx>,
/// The parent query job which created this job and is implicitly waiting on it.
pub parent: Option<Lrc<QueryJob<'tcx>>>,
2018-03-24 00:19:20 -05:00
/// Diagnostic messages which are emitted while the query executes
pub diagnostics: Lock<Vec<Diagnostic>>,
#[cfg(parallel_queries)]
2018-05-27 02:01:57 -05:00
latch: QueryLatch<'tcx>,
}
impl<'tcx> QueryJob<'tcx> {
2018-03-24 00:19:20 -05:00
/// Creates a new query job
pub fn new(info: QueryInfo<'tcx>, parent: Option<Lrc<QueryJob<'tcx>>>) -> Self {
QueryJob {
diagnostics: Lock::new(Vec::new()),
2018-03-24 00:19:20 -05:00
info,
parent,
#[cfg(parallel_queries)]
latch: QueryLatch::new(),
}
}
2018-03-24 00:19:20 -05:00
/// Awaits for the query job to complete.
///
/// For single threaded rustc there's no concurrent jobs running, so if we are waiting for any
/// query that means that there is a query cycle, thus this always running a cycle error.
pub(super) fn await<'lcx>(
&self,
tcx: TyCtxt<'_, 'tcx, 'lcx>,
span: Span,
) -> Result<(), CycleError<'tcx>> {
#[cfg(not(parallel_queries))]
{
self.find_cycle_in_stack(tcx, span)
}
#[cfg(parallel_queries)]
{
tls::with_related_context(tcx, move |icx| {
let mut waiter = QueryWaiter {
query: &icx.query,
span,
cycle: None,
condvar: Condvar::new(),
};
self.latch.await(&mut waiter);
match waiter.cycle {
None => Ok(()),
Some(cycle) => Err(cycle)
}
})
}
}
#[cfg(not(parallel_queries))]
fn find_cycle_in_stack<'lcx>(
&self,
tcx: TyCtxt<'_, 'tcx, 'lcx>,
span: Span,
) -> Result<(), CycleError<'tcx>> {
2018-03-24 00:19:20 -05:00
// Get the current executing query (waiter) and find the waitee amongst its parents
let mut current_job = tls::with_related_context(tcx, |icx| icx.query.clone());
let mut cycle = Vec::new();
while let Some(job) = current_job {
2018-03-24 00:19:20 -05:00
cycle.insert(0, job.info.clone());
if &*job as *const _ == self as *const _ {
2018-04-13 15:20:10 -05:00
// This is the end of the cycle
// The span entry we included was for the usage
// of the cycle itself, and not part of the cycle
// Replace it with the span which caused the cycle to form
cycle[0].span = span;
// Find out why the cycle itself was used
let usage = job.parent.as_ref().map(|parent| {
(job.info.span, parent.info.query.clone())
});
return Err(CycleError { usage, cycle });
}
current_job = job.parent.clone();
}
2018-04-13 15:20:10 -05:00
panic!("did not find a cycle")
}
2018-03-24 00:19:20 -05:00
/// Signals to waiters that the query is complete.
///
/// This does nothing for single threaded rustc,
/// as there are no concurrent jobs which could be waiting on us
2018-05-27 02:01:57 -05:00
pub fn signal_complete(&self) {
#[cfg(parallel_queries)]
2018-05-27 02:01:57 -05:00
self.latch.set();
}
}
#[cfg(parallel_queries)]
2018-05-27 02:01:57 -05:00
struct QueryWaiter<'tcx> {
query: *const Option<Lrc<QueryJob<'tcx>>>,
condvar: Condvar,
span: Span,
cycle: Option<CycleError<'tcx>>,
}
#[cfg(parallel_queries)]
2018-05-27 02:01:57 -05:00
impl<'tcx> QueryWaiter<'tcx> {
fn notify(&self, registry: &rayon_core::Registry) {
rayon_core::mark_unblocked(registry);
self.condvar.notify_one();
}
}
#[cfg(parallel_queries)]
2018-05-27 02:01:57 -05:00
struct QueryLatchInfo<'tcx> {
complete: bool,
2018-05-27 02:01:57 -05:00
waiters: Vec<*mut QueryWaiter<'tcx>>,
}
2018-05-27 02:01:57 -05:00
// Required because of raw pointers
#[cfg(parallel_queries)]
2018-05-27 02:01:57 -05:00
unsafe impl<'tcx> Send for QueryLatchInfo<'tcx> {}
#[cfg(parallel_queries)]
struct QueryLatch<'tcx> {
info: Mutex<QueryLatchInfo<'tcx>>,
}
#[cfg(parallel_queries)]
2018-05-27 02:01:57 -05:00
impl<'tcx> QueryLatch<'tcx> {
fn new() -> Self {
QueryLatch {
info: Mutex::new(QueryLatchInfo {
complete: false,
waiters: Vec::new(),
}),
}
}
2018-05-27 06:05:41 -05:00
/// Awaits the caller on this latch by blocking the current thread.
2018-05-27 02:01:57 -05:00
fn await(&self, waiter: &mut QueryWaiter<'tcx>) {
let mut info = self.info.lock();
if !info.complete {
2018-05-27 06:05:41 -05:00
// We push the waiter on to the `waiters` list. It can be accessed inside
// the `wait` call below, by 1) the `set` method or 2) by deadlock detection.
// Both of these will remove it from the `waiters` list before resuming
// this thread.
2018-05-27 02:01:57 -05:00
info.waiters.push(waiter);
2018-05-27 06:05:41 -05:00
// If this detects a deadlock and the deadlock handler want to resume this thread
// we have to be in the `wait` call. This is ensured by the deadlock handler
// getting the self.info lock.
rayon_core::mark_blocked();
2018-05-27 06:05:41 -05:00
waiter.condvar.wait(&mut info);
}
}
2018-05-27 06:05:41 -05:00
/// Sets the latch and resumes all waiters on it
2018-05-27 02:01:57 -05:00
fn set(&self) {
let mut info = self.info.lock();
debug_assert!(!info.complete);
info.complete = true;
let registry = rayon_core::Registry::current();
for waiter in info.waiters.drain(..) {
2018-05-27 02:01:57 -05:00
unsafe {
(*waiter).notify(&registry);
}
}
}
2018-05-27 06:05:41 -05:00
/// Remove a single waiter from the list of waiters.
/// This is used to break query cycles.
fn extract_waiter(
&self,
waiter: usize,
2018-05-27 02:01:57 -05:00
) -> *mut QueryWaiter<'tcx> {
let mut info = self.info.lock();
debug_assert!(!info.complete);
// Remove the waiter from the list of waiters
2018-05-27 06:05:41 -05:00
info.waiters.remove(waiter)
}
}
2018-05-27 06:05:41 -05:00
/// A pointer to an active query job. This is used to give query jobs an identity.
#[cfg(parallel_queries)]
type Ref<'tcx> = *const QueryJob<'tcx>;
2018-05-27 06:05:41 -05:00
/// A resumable waiter of a query. The usize is the index into waiters in the query's latch
#[cfg(parallel_queries)]
type Waiter<'tcx> = (Ref<'tcx>, usize);
2018-05-27 06:05:41 -05:00
/// Visits all the non-resumable and resumable waiters of a query.
/// Only waiters in a query are visited.
/// `visit` is called for every waiter and is passed a query waiting on `query_ref`
/// and a span indicating the reason the query waited on `query_ref`.
/// If `visit` returns Some, this function returns.
/// For visits of non-resumable waiters it returns the return value of `visit`.
/// For visits of resumable waiters it returns Some(Some(Waiter)) which has the
/// required information to resume the waiter.
/// If all `visit` calls returns None, this function also returns None.
#[cfg(parallel_queries)]
fn visit_waiters<'tcx, F>(query_ref: Ref<'tcx>, mut visit: F) -> Option<Option<Waiter<'tcx>>>
where
F: FnMut(Span, Ref<'tcx>) -> Option<Option<Waiter<'tcx>>>
{
let query = unsafe { &*query_ref };
2018-05-27 06:05:41 -05:00
// Visit the parent query which is a non-resumable waiter since it's on the same stack
if let Some(ref parent) = query.parent {
if let Some(cycle) = visit(query.info.span, &**parent as Ref) {
return Some(cycle);
}
}
2018-05-27 06:05:41 -05:00
// Visit the explict waiters which use condvars and are resumable
2018-05-27 02:01:57 -05:00
for (i, &waiter) in query.latch.info.lock().waiters.iter().enumerate() {
unsafe {
if let Some(ref waiter_query) = *(*waiter).query {
if visit((*waiter).span, &**waiter_query as Ref).is_some() {
2018-05-27 06:05:41 -05:00
// Return a value which indicates that this waiter can be resumed
2018-05-27 02:01:57 -05:00
return Some(Some((query_ref, i)));
}
}
}
}
None
}
2018-05-27 06:05:41 -05:00
/// Look for query cycles by doing a depth first search starting at `query`.
/// `span` is the reason for the `query` to execute. This is initially DUMMY_SP.
/// If a cycle is detected, this initial value is replaced with the span causing
/// the cycle.
#[cfg(parallel_queries)]
fn cycle_check<'tcx>(query: Ref<'tcx>,
span: Span,
stack: &mut Vec<(Span, Ref<'tcx>)>,
visited: &mut HashSet<Ref<'tcx>>) -> Option<Option<Waiter<'tcx>>> {
if visited.contains(&query) {
return if let Some(p) = stack.iter().position(|q| q.1 == query) {
2018-05-27 06:05:41 -05:00
// We detected a query cycle, fix up the initial span and return Some
// Remove previous stack entries
stack.splice(0..p, iter::empty());
// Replace the span for the first query with the cycle cause
stack[0].0 = span;
Some(None)
} else {
None
}
}
2018-05-27 06:05:41 -05:00
// Mark this query is visited and add it to the stack
visited.insert(query);
stack.push((span, query));
2018-05-27 06:05:41 -05:00
// Visit all the waiters
let r = visit_waiters(query, |span, successor| {
cycle_check(successor, span, stack, visited)
});
2018-05-27 06:05:41 -05:00
// Remove the entry in our stack if we didn't find a cycle
if r.is_none() {
stack.pop();
}
r
}
2018-05-27 06:05:41 -05:00
/// Finds out if there's a path to the compiler root (aka. code which isn't in a query)
/// from `query` without going through any of the queries in `visited`.
/// This is achieved with a depth first search.
#[cfg(parallel_queries)]
fn connected_to_root<'tcx>(query: Ref<'tcx>, visited: &mut HashSet<Ref<'tcx>>) -> bool {
2018-05-27 06:05:41 -05:00
// We already visited this or we're deliberately ignoring it
if visited.contains(&query) {
return false;
}
2018-05-27 06:05:41 -05:00
// This query is connected to the root (it has no query parent), return true
if unsafe { (*query).parent.is_none() } {
return true;
}
visited.insert(query);
let mut connected = false;
visit_waiters(query, |_, successor| {
if connected_to_root(successor, visited) {
Some(None)
} else {
None
}
}).is_some()
}
2018-05-27 06:05:41 -05:00
/// Looks for query cycles starting from the last query in `jobs`.
/// If a cycle is found, all queries in the cycle is removed from `jobs` and
/// the function return true.
/// If a cycle was not found, the starting query is removed from `jobs` and
/// the function returns false.
#[cfg(parallel_queries)]
fn remove_cycle<'tcx>(
jobs: &mut Vec<Ref<'tcx>>,
2018-05-27 02:01:57 -05:00
wakelist: &mut Vec<*mut QueryWaiter<'tcx>>,
tcx: TyCtxt<'_, 'tcx, '_>
2018-05-27 06:05:41 -05:00
) -> bool {
let mut visited = HashSet::new();
let mut stack = Vec::new();
2018-05-27 06:05:41 -05:00
// Look for a cycle starting with the last query in `jobs`
if let Some(waiter) = cycle_check(jobs.pop().unwrap(),
DUMMY_SP,
&mut stack,
&mut visited) {
// Reverse the stack so earlier entries require later entries
stack.reverse();
2018-05-27 06:05:41 -05:00
// Extract the spans and queries into separate arrays
let mut spans: Vec<_> = stack.iter().map(|e| e.0).collect();
let queries = stack.iter().map(|e| e.1);
2018-05-27 06:05:41 -05:00
// Shift the spans so that queries are matched with the span for their waitee
let last = spans.pop().unwrap();
spans.insert(0, last);
2018-05-27 06:05:41 -05:00
// Zip them back together
let mut stack: Vec<_> = spans.into_iter().zip(queries).collect();
// Remove the queries in our cycle from the list of jobs to look at
for r in &stack {
jobs.remove_item(&r.1);
}
// Find the queries in the cycle which are
// connected to queries outside the cycle
let entry_points: Vec<Ref<'_>> = stack.iter().filter_map(|query| {
// Mark all the other queries in the cycle as already visited
let mut visited = HashSet::from_iter(stack.iter().filter_map(|q| {
if q.1 != query.1 {
Some(q.1)
} else {
None
}
}));
if connected_to_root(query.1, &mut visited) {
Some(query.1)
} else {
None
}
}).collect();
// Deterministically pick an entry point
// FIXME: Sort this instead
let mut hcx = tcx.create_stable_hashing_context();
let entry_point = *entry_points.iter().min_by_key(|&&q| {
let mut stable_hasher = StableHasher::<u64>::new();
unsafe { (*q).info.query.hash_stable(&mut hcx, &mut stable_hasher); }
stable_hasher.finish()
}).unwrap();
// Shift the stack until our entry point is first
while stack[0].1 != entry_point {
let last = stack.pop().unwrap();
stack.insert(0, last);
}
2018-05-27 06:05:41 -05:00
// Create the cycle error
let mut error = CycleError {
usage: None,
cycle: stack.iter().map(|&(s, q)| QueryInfo {
span: s,
query: unsafe { (*q).info.query.clone() },
} ).collect(),
};
2018-05-27 06:05:41 -05:00
// We unwrap `waiter` here since there must always be one
// edge which is resumeable / waited using a query latch
let (waitee_query, waiter_idx) = waiter.unwrap();
let waitee_query = unsafe { &*waitee_query };
// Extract the waiter we want to resume
let waiter = waitee_query.latch.extract_waiter(waiter_idx);
// Set the cycle error it will be picked it up when resumed
unsafe {
(*waiter).cycle = Some(error);
}
// Put the waiter on the list of things to resume
wakelist.push(waiter);
true
} else {
false
}
}
2018-05-27 06:05:41 -05:00
/// Creates a new thread and forwards information in thread locals to it.
/// The new thread runs the deadlock handler.
#[cfg(parallel_queries)]
pub fn handle_deadlock() {
use syntax;
use syntax_pos;
let registry = rayon_core::Registry::current();
let gcx_ptr = tls::GCX_PTR.with(|gcx_ptr| {
gcx_ptr as *const _
});
let gcx_ptr = unsafe { &*gcx_ptr };
let syntax_globals = syntax::GLOBALS.with(|syntax_globals| {
syntax_globals as *const _
});
let syntax_globals = unsafe { &*syntax_globals };
let syntax_pos_globals = syntax_pos::GLOBALS.with(|syntax_pos_globals| {
syntax_pos_globals as *const _
});
let syntax_pos_globals = unsafe { &*syntax_pos_globals };
thread::spawn(move || {
tls::GCX_PTR.set(gcx_ptr, || {
syntax_pos::GLOBALS.set(syntax_pos_globals, || {
syntax_pos::GLOBALS.set(syntax_pos_globals, || {
tls::with_thread_locals(|| {
unsafe {
tls::with_global(|tcx| deadlock(tcx, &registry))
}
})
})
})
})
});
}
2018-05-27 06:05:41 -05:00
/// Detects query cycles by using depth first search over all active query jobs.
/// If a query cycle is found it will break the cycle by finding an edge which
/// uses a query latch and then resuming that waiter.
/// There may be multiple cycles involved in a deadlock, so this searches
/// all active queries for cycles before finally resuming all the waiters at once.
#[cfg(parallel_queries)]
fn deadlock(tcx: TyCtxt<'_, '_, '_>, registry: &rayon_core::Registry) {
let on_panic = OnDrop(|| {
eprintln!("deadlock handler panicked, aborting process");
process::abort();
});
let mut wakelist = Vec::new();
let mut jobs: Vec<_> = tcx.maps.collect_active_jobs().iter().map(|j| &**j as Ref).collect();
2018-05-27 06:05:41 -05:00
let mut found_cycle = false;
while jobs.len() > 0 {
2018-05-27 06:05:41 -05:00
if remove_cycle(&mut jobs, &mut wakelist, tcx) {
found_cycle = true;
}
}
2018-05-27 06:05:41 -05:00
// Check that a cycle was found. It is possible for a deadlock to occur without
// a query cycle if a query which can be waited on uses Rayon to do multithreading
// internally. Such a query (X) may be executing on 2 threads (A and B) and A may
// wait using Rayon on B. Rayon may then switch to executing another query (Y)
// which in turn will wait on X causing a deadlock. We have a false dependency from
// X to Y due to Rayon waiting and a true dependency from Y to X. The algorithm here
// only considers the true dependency and won't detect a cycle.
assert!(found_cycle);
// FIXME: Ensure this won't cause a deadlock before we return
for waiter in wakelist.into_iter() {
2018-05-27 02:01:57 -05:00
unsafe {
(*waiter).notify(registry);
}
}
2018-05-27 00:47:44 -05:00
on_panic.disable();
}