Skip to content

New parallel expression in Q# #3298

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
swernli wants to merge 15 commits into
base: main
Choose a base branch
from
Open

New parallel expression in Q# #3298

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
15 commits
Select commit Hold shift + click to select a range
2071da6
New `parallel` expression
swernli Apr 9, 2026
4228eda
Add tests for parallel and parallel within expressions
swernli May 13, 2026
732d057
Add qsc_partial_eval tests for parallel and parallel within expressions
swernli May 14, 2026
1f2738e
track limits including allocated qubits
swernli May 28, 2026
734cf9f
Remove incorrectly added egg info
swernli Jun 8, 2026
5b53d95
Merge remote-tracking branch 'origin/main' into swernli/parallel
swernli Jun 12, 2026
dc711b5
FIR pass updates
swernli Jun 12, 2026
e880df1
Add FIR transform tests for parallel, fix handling
swernli Jun 13, 2026
f237352
Always lower parallel into a block
swernli Jun 17, 2026
a1d8b0e
Merge remote-tracking branch 'origin/main' into swernli/parallel
swernli Jun 18, 2026
5f1d604
Merge remote-tracking branch 'origin/main' into swernli/parallel
swernli Jun 29, 2026
db764c9
Fix tests
swernli Jun 29, 2026
bcbaab2
Update HIR display
swernli Jun 29, 2026
bd1e204
Update AST display and enum form
swernli Jun 29, 2026
39a4783
Use enum instead of bool
swernli Jun 29, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
229 changes: 229 additions & 0 deletions source/compiler/qsc/src/interpret/circuit_tests.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2137,3 +2137,232 @@ mod debugger_stepping {
.assert_eq(&circs);
}
}

// Without parallel, released qubits have their IDs recycled on subsequent allocations.
// The inner block releases q1/q2, so q3/q4 reuse the same wires (q_0, q_1).
#[test]
fn parallel_baseline_qubit_ids_recycled_without_parallel() {
let circ = circuit_without_groups(
r"
namespace Test {
@EntryPoint()
operation Main() : Unit {
{ use q1 = Qubit(); H(q1); use q2 = Qubit(); H(q2); }
use q3 = Qubit();
H(q3);
use q4 = Qubit();
H(q4);
}
}
",
CircuitEntryPoint::EntryPoint,
);

expect![[r#"
q_0@test.qs:4:22, test.qs:5:20 ─ H@test.qs:4:40 ─── H@test.qs:6:20 ──
q_1@test.qs:4:47, test.qs:7:20 ─ H@test.qs:4:65 ─── H@test.qs:8:20 ──
"#]]
.assert_eq(&circ);
}

// Inside a parallel expression all releases are deferred, so q3/q4 get fresh wires
// instead of reusing q_0/q_1. This mirrors the baseline test with `parallel` added.
#[test]
fn parallel_defers_qubit_release() {
let circ = circuit_without_groups(
r"
namespace Test {
@EntryPoint()
operation Main() : Unit {
parallel {
{ use q1 = Qubit(); H(q1); use q2 = Qubit(); H(q2); }
use q3 = Qubit();
H(q3);
use q4 = Qubit();
H(q4);
}
}
}
",
CircuitEntryPoint::EntryPoint,
);

expect![[r#"
q_0@test.qs:5:26 ─ H@test.qs:5:44 ──
q_1@test.qs:5:51 ─ H@test.qs:5:69 ──
q_2@test.qs:6:24 ─ H@test.qs:7:24 ──
q_3@test.qs:8:24 ─ H@test.qs:9:24 ──
"#]]
.assert_eq(&circ);
}

// After a parallel block ends its deferred releases become available, so a second
// parallel block reuses the same qubit wires.
#[test]
fn parallel_releases_available_after_block_ends() {
let circ = circuit_without_groups(
r"
namespace Test {
@EntryPoint()
operation Main() : Unit {
parallel {
use q = Qubit();
H(q);
}
parallel {
use q = Qubit();
X(q);
}
}
}
",
CircuitEntryPoint::EntryPoint,
);

expect![[r#"
q_0@test.qs:5:24, test.qs:9:24 ─ H@test.qs:6:24 ─── X@test.qs:10:24 ─
"#]]
.assert_eq(&circ);
}

// In nested parallel expressions, inner block qubits flow to the outer layer on removal
// so the outer block allocates fresh wires even after the inner block ends.
#[test]
fn parallel_nested_defers_inner_releases_to_outer() {
let circ = circuit_without_groups(
r"
namespace Test {
@EntryPoint()
operation Main() : Unit {
parallel {
use outer = Qubit();
H(outer);
parallel {
use inner1 = Qubit();
H(inner1);
use inner2 = Qubit();
H(inner2);
}
use outer2 = Qubit();
H(outer2);
}
}
}
",
CircuitEntryPoint::EntryPoint,
);

expect![[r#"
q_0@test.qs:5:24 ─ H@test.qs:6:24 ──
q_1@test.qs:8:28 ─ H@test.qs:9:28 ──
q_2@test.qs:10:28 ─ H@test.qs:11:28 ─
q_3@test.qs:13:24 ─ H@test.qs:14:24 ─
"#]]
.assert_eq(&circ);
}

// parallel within N: once N qubits are deferred the pool replenishes, so later
// allocations reuse existing wires rather than creating new ones.
#[test]
fn parallel_within_reuses_wires_after_limit() {
let circ = circuit_without_groups(
r"
namespace Test {
@EntryPoint()
operation Main() : Unit {
parallel within 2 {
{ use q1 = Qubit(); H(q1); }
{ use q2 = Qubit(); H(q2); }
{ use q3 = Qubit(); H(q3); }
{ use q4 = Qubit(); H(q4); }
}
}
}
",
CircuitEntryPoint::EntryPoint,
);

expect![[r#"
q_0@test.qs:5:26 ─ H@test.qs:5:44 ─── H@test.qs:7:44 ──
q_1@test.qs:6:26 ─ H@test.qs:6:44 ─── H@test.qs:8:44 ──
"#]]
.assert_eq(&circ);
}

// Outer `parallel within 6` with inner `parallel within 2`. The inner limit reuses
// wires within each iteration. Once the outer deferred count reaches 6 (iteration 3),
// the outer layer replenishes and reuses its wires too.
#[test]
fn parallel_within_nested_defers_through_outer_limit() {
let circ = circuit_without_groups(
r"
namespace Test {
@EntryPoint()
operation Main() : Unit {
parallel within 6 {
for _ in 0..2 {
{ use q0 = Qubit(); H(q0); }
parallel within 2 {
{ use q1 = Qubit(); H(q1); }
{ use q2 = Qubit(); H(q2); }
{ use q3 = Qubit(); H(q3); }
{ use q4 = Qubit(); H(q4); }
}
}
}
}
}
",
CircuitEntryPoint::EntryPoint,
);

expect![[r#"

@ScottCarda-MS ScottCarda-MS Jun 18, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It is difficult for me to understand what should be expected from the nested parallel within expressions, so I'll double-check with you. Is this correct and what is expected with the nested setup?

q_0@test.qs:6:30 ─ H@test.qs:6:48 ─── H@test.qs:6:48 ────────────────────────────────────────
q_1@test.qs:8:34 ─ H@test.qs:8:52 ─── H@test.qs:10:52 ── H@test.qs:8:52 ─── H@test.qs:10:52 ─
q_2@test.qs:9:34 ─ H@test.qs:9:52 ─── H@test.qs:11:52 ── H@test.qs:9:52 ─── H@test.qs:11:52 ─
q_3@test.qs:6:30 ─ H@test.qs:6:48 ───────────────────────────────────────────────────────────
q_4@test.qs:8:34 ─ H@test.qs:8:52 ─── H@test.qs:10:52 ───────────────────────────────────────
q_5@test.qs:9:34 ─ H@test.qs:9:52 ─── H@test.qs:11:52 ───────────────────────────────────────
"#]].assert_eq(&circ);
}

// Same structure but the outer parallel has no limit. The inner `parallel within 2`
// still reuses within each iteration, but the outer unlimited layer never replenishes,
// so iteration 3 allocates fresh wires (q_6/q_7/q_8) instead of reusing q_0/q_1/q_2.
#[test]
fn parallel_nested_unlimited_outer_defers_all() {
let circ = circuit_without_groups(
r"
namespace Test {
@EntryPoint()
operation Main() : Unit {
parallel {
for _ in 0..2 {
{ use q0 = Qubit(); H(q0); }
parallel within 2 {
{ use q1 = Qubit(); H(q1); }
{ use q2 = Qubit(); H(q2); }
{ use q3 = Qubit(); H(q3); }
{ use q4 = Qubit(); H(q4); }
}
}
}
}
}
",
CircuitEntryPoint::EntryPoint,
);

expect![[r#"
q_0@test.qs:6:30 ─ H@test.qs:6:48 ─────────────────────
q_1@test.qs:8:34 ─ H@test.qs:8:52 ─── H@test.qs:10:52 ─
q_2@test.qs:9:34 ─ H@test.qs:9:52 ─── H@test.qs:11:52 ─
q_3@test.qs:6:30 ─ H@test.qs:6:48 ─────────────────────
q_4@test.qs:8:34 ─ H@test.qs:8:52 ─── H@test.qs:10:52 ─
q_5@test.qs:9:34 ─ H@test.qs:9:52 ─── H@test.qs:11:52 ─
q_6@test.qs:6:30 ─ H@test.qs:6:48 ─────────────────────
q_7@test.qs:8:34 ─ H@test.qs:8:52 ─── H@test.qs:10:52 ─
q_8@test.qs:9:34 ─ H@test.qs:9:52 ─── H@test.qs:11:52 ─
"#]]
.assert_eq(&circ);
}
17 changes: 17 additions & 0 deletions source/compiler/qsc_ast/src/ast.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -906,6 +906,8 @@ pub enum ExprKind {
Lambda(CallableKind, Box<Pat>, Box<Expr>),
/// A literal.
Lit(Box<Lit>),
/// A parallel expression: `parallel a`, optionally including a limit: `parallel within n a`
Parallel(Option<Box<Expr>>, Box<Expr>),
/// Parentheses: `(a)`.
Paren(Box<Expr>),
/// A path: `a` or `a.b`.
Expand Down Expand Up @@ -953,6 +955,7 @@ impl Display for ExprKind {
ExprKind::Interpolate(components) => display_interpolate(indent, components)?,
ExprKind::Lambda(kind, param, expr) => display_lambda(indent, *kind, param, expr)?,
ExprKind::Lit(lit) => write!(indent, "Lit: {lit}")?,
ExprKind::Parallel(limit, body) => display_parallel(indent, limit.as_deref(), body)?,
ExprKind::Paren(e) => write!(indent, "Paren: {e}")?,
ExprKind::Path(p) => write!(indent, "Path: {p}")?,
ExprKind::Range(start, step, end) => {
Expand Down Expand Up @@ -1126,6 +1129,20 @@ fn display_interpolate(
Ok(())
}

fn display_parallel(
mut indent: Indented<Formatter>,
limit: Option<&Expr>,
body: &Expr,
) -> fmt::Result {
write!(indent, "Parallel:")?;
indent = set_indentation(indent, 1);
if let Some(l) = limit {
write!(indent, "\nLimit: {l}")?;
}
write!(indent, "\nBody: {body}")?;
Ok(())
}

fn display_lambda(
mut indent: Indented<Formatter>,
kind: CallableKind,
Expand Down
6 changes: 6 additions & 0 deletions source/compiler/qsc_ast/src/mut_visit.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -365,6 +365,12 @@ pub fn walk_expr(vis: &mut impl MutVisitor, expr: &mut Expr) {
ExprKind::Paren(expr) | ExprKind::Return(expr) | ExprKind::UnOp(_, expr) => {
vis.visit_expr(expr);
}
ExprKind::Parallel(limit, body) => {
if let Some(l) = limit.as_mut() {
vis.visit_expr(l);
}
vis.visit_expr(body);
}
ExprKind::Path(path) => vis.visit_path_kind(path),
ExprKind::Range(start, step, end) => {
for s in start.iter_mut() {
Expand Down
6 changes: 6 additions & 0 deletions source/compiler/qsc_ast/src/visit.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -337,6 +337,12 @@ pub fn walk_expr<'a>(vis: &mut impl Visitor<'a>, expr: &'a Expr) {
ExprKind::Paren(expr) | ExprKind::Return(expr) | ExprKind::UnOp(_, expr) => {
vis.visit_expr(expr);
}
ExprKind::Parallel(limit, body) => {
if let Some(l) = limit.as_ref() {
vis.visit_expr(l);
}
vis.visit_expr(body);
}
ExprKind::Path(path) => vis.visit_path_kind(path),
ExprKind::Range(start, step, end) => {
if let Some(s) = start.as_ref() {
Expand Down
9 changes: 9 additions & 0 deletions source/compiler/qsc_codegen/src/qsharp.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -540,6 +540,15 @@ impl<W: Write> Visitor<'_> for QSharpGen<W> {
}
self.visit_expr(expr);
}
ExprKind::Parallel(limit, body) => {
self.write("parallel ");
if let Some(limit) = limit {
self.write("within ");
self.visit_expr(limit);
self.write(" ");
}
self.visit_expr(body);
}
ExprKind::Paren(expr) => {
self.write("(");
self.visit_expr(expr);
Expand Down
Loading
Loading