Skip to content

Commit

Permalink
fix: erratic stack overflow in infer.rs (live_var) (#638)
Browse files Browse the repository at this point in the history
closes #633

* Add test case. Reduced this to a 2-node DFG (plus Input and Output).
That this hasn't been showing up repeatedly in other tests - I put down
to #388: most custom ops have empty `extension_reqs` so do not generate
Plus constraints.
* Run (50%-failing) test 10 times -> failure rate ~~ 1 in 2^10
* Fix calculation of `live_vars` and `live_metas` by one-off traversal
of constraints+solutions in `fn results()`.

---------

Co-authored-by: Craig Roy <[email protected]>
  • Loading branch information
acl-cqc and croyzor authored Nov 7, 2023
1 parent 57270d5 commit b7ebb0d
Showing 1 changed file with 86 additions and 44 deletions.
130 changes: 86 additions & 44 deletions src/extension/infer.rs
Original file line number Diff line number Diff line change
Expand Up @@ -22,7 +22,7 @@ use super::validate::ExtensionError;

use petgraph::graph as pg;

use std::collections::{HashMap, HashSet};
use std::collections::{HashMap, HashSet, VecDeque};

use thiserror::Error;

Expand Down Expand Up @@ -521,57 +521,46 @@ impl UnificationContext {
pub fn results(&self) -> Result<ExtensionSolution, InferExtensionError> {
// Check that all of the metavariables associated with nodes of the
// graph are solved
let depended_upon = {
let mut h: HashMap<Meta, Vec<Meta>> = HashMap::new();
for (m, m2) in self.constraints.iter().flat_map(|(m, cs)| {
cs.iter().flat_map(|c| match c {
Constraint::Plus(_, m2) => Some((*m, self.resolve(*m2))),
_ => None,
})
}) {
h.entry(m2).or_default().push(m);
}
h
};
// Calculate everything dependent upon a variable.
// Note it would be better to find metas ALL of whose dependencies were (transitively)
// on variables, but this is more complex, and hard to define if there are cycles
// of PLUS constraints, so leaving that as a TODO until we've handled such cycles.
let mut depends_on_var = HashSet::new();
let mut queue = VecDeque::from_iter(self.variables.iter());
while let Some(m) = queue.pop_front() {
if depends_on_var.insert(m) {
if let Some(d) = depended_upon.get(m) {
queue.extend(d.iter())
}
}
}

let mut results: ExtensionSolution = HashMap::new();
for (loc, meta) in self.extensions.iter() {
if let Some(rs) = self.get_solution(meta) {
if loc.1 == Direction::Incoming {
results.insert(loc.0, rs.clone());
}
} else if self.live_var(meta).is_some() {
// If it depends on some other live meta, that's bad news.
return Err(InferExtensionError::Unsolved { location: *loc });
}
// If it only depends on graph variables, then we don't have
// a *solution*, but it's fine
}
debug_assert!(self.live_metas().is_empty());
Ok(results)
}

// Get the live var associated with a meta.
// TODO: This should really be a list
fn live_var(&self, m: &Meta) -> Option<Meta> {
if self.variables.contains(m) || self.variables.contains(&self.resolve(*m)) {
return None;
}

// TODO: We should be doing something to ensure that these are the same check...
if self.get_solution(m).is_none() {
if let Some(cs) = self.get_constraints(m) {
for c in cs {
match c {
Constraint::Plus(_, m) => return self.live_var(m),
_ => panic!("we shouldn't be here!"),
}
} else {
// Unsolved nodes must be unsolved because they depend on graph variables.
if !depends_on_var.contains(&self.resolve(*meta)) {
return Err(InferExtensionError::Unsolved { location: *loc });
}
}
Some(*m)
} else {
None
}
}

/// Return the set of "live" metavariables in the context.
/// "Live" here means a metavariable:
/// - Is associated to a location in the graph in `UnifyContext.extensions`
/// - Is still unsolved
/// - Isn't a variable
fn live_metas(&self) -> HashSet<Meta> {
self.extensions
.values()
.filter_map(|m| self.live_var(m))
.filter(|m| !self.variables.contains(m))
.collect()
Ok(results)
}

/// Iterates over a set of metas (the argument) and tries to solve
Expand Down Expand Up @@ -665,12 +654,16 @@ mod test {

use super::*;
use crate::builder::test::closed_dfg_root_hugr;
use crate::builder::{DFGBuilder, Dataflow, DataflowHugr};
use crate::extension::prelude::QB_T;
use crate::extension::ExtensionId;
use crate::extension::{prelude::PRELUDE_REGISTRY, ExtensionSet};
use crate::hugr::{validate::ValidationError, Hugr, HugrMut, HugrView, NodeType};
use crate::macros::const_extension_ids;
use crate::ops::OpType;
use crate::ops::custom::{ExternalOp, OpaqueOp};
use crate::ops::{self, dataflow::IOTrait, handle::NodeHandle, OpTrait};
use crate::ops::{LeafOp, OpType};

use crate::type_row;
use crate::types::{FunctionType, Type, TypeRow};

Expand Down Expand Up @@ -1539,4 +1532,53 @@ mod test {

Ok(())
}

/// This was stack-overflowing approx 50% of the time,
/// see https://github.com/CQCL/hugr/issues/633
#[test]
fn plus_on_self() -> Result<(), Box<dyn std::error::Error>> {
let ext = ExtensionId::new("unknown1").unwrap();
let delta = ExtensionSet::singleton(&ext);
let ft = FunctionType::new_linear(type_row![QB_T, QB_T]).with_extension_delta(&delta);
let mut dfg = DFGBuilder::new(ft.clone())?;

// While https://github.com/CQCL-DEV/hugr/issues/388 is unsolved,
// most operations have empty extension_reqs (not including their own extension).
// Define some that do.
let binop: LeafOp = ExternalOp::Opaque(OpaqueOp::new(
ext.clone(),
"2qb_op",
String::new(),
vec![],
Some(ft),
))
.into();
let unary_sig = FunctionType::new_linear(type_row![QB_T])
.with_extension_delta(&ExtensionSet::singleton(&ext));
let unop: LeafOp = ExternalOp::Opaque(OpaqueOp::new(
ext,
"1qb_op",
String::new(),
vec![],
Some(unary_sig),
))
.into();
// Constrain q1,q2 as PLUS(ext1, inputs):
let [q1, q2] = dfg
.add_dataflow_op(binop.clone(), dfg.input_wires())?
.outputs_arr();
// Constrain q1 as PLUS(ext2, q2):
let [q1] = dfg.add_dataflow_op(unop, [q1])?.outputs_arr();
// Constrain q1 as EQUALS(q2) by using both together
dfg.finish_hugr_with_outputs([q1, q2], &PRELUDE_REGISTRY)?;
// The combined q1+q2 variable now has two PLUS constraints - on itself and the inputs.
Ok(())
}

/// [plus_on_self] had about a 50% rate of failing with stack overflow.
/// So if we run 10 times, that would succeed about 1 run in 2^10, i.e. <0.1%
#[test]
fn plus_on_self_10_times() {
[0; 10].iter().for_each(|_| plus_on_self().unwrap())
}
}

0 comments on commit b7ebb0d

Please sign in to comment.