✨ Add SWAPAbsorption Pass by MatthiasReumann · Pull Request #1683 · munich-quantum-toolkit/core

MatthiasReumann · 2026-05-04T07:14:38Z

Description

This pull request adds the SWAPAbsorption pass, which absorbs SWAPs by exchanging static qubits.

Checklist

The pull request only contains commits that are focused and relevant to this change.
I have added appropriate tests that cover the new/changed functionality.
I have updated the documentation to reflect these changes.
I have added entries to the changelog for any noteworthy additions, changes, fixes, or removals.
I have added migration instructions to the upgrade guide (if needed).
The changes follow the project's style guidelines and introduce no new warnings.
The changes are fully tested and pass the CI checks.
I have reviewed my own code changes.

If PR contains AI-assisted content:

I have disclosed the use of AI tools in the PR description as per our AI Usage Guidelines.
AI-assisted commits include an Assisted-by: [Model Name] via [Tool Name] footer.
I confirm that I have personally reviewed and understood all AI-generated content, and accept full responsibility for it.

codecov · 2026-05-04T07:19:23Z

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

MatthiasReumann

@jmoosburger

As discussed in the previous pull request, a few comments on how to proceed with this task.

I think you should be able to sync your fork and checkout this branch. If not, just let me know! 🙋🏻‍♂️

Thanks for working on this 🙏

MatthiasReumann · 2026-05-04T07:19:45Z

+TEST_F(SwapAbsorbPassTest, PassDoesNotChangeSwaplessProgram) {
+
+  qco::QCOProgramBuilder builder(context.get());
+  builder.initialize();
+
+  const auto q00 = builder.allocQubit();
+  const auto q10 = builder.allocQubit();
+  const auto q20 = builder.allocQubit();
+
+  const auto q01 = builder.h(q00);
+  const auto [q02, q11] = builder.cx(q01, q10);
+
+  builder.sink(q02);
+  builder.sink(q11);
+
+  auto moduleThroughPass = builder.finalize();
+  auto originalModule = moduleThroughPass->clone();
+
+  applySwapAbsorb(moduleThroughPass);
+  ASSERT_TRUE(mlir::OperationEquivalence::isEquivalentTo(
+      moduleThroughPass.get(), originalModule,
+      mlir::OperationEquivalence::Flags::IgnoreLocations));
+}


Because there is also a QCO program builder, we don't require the conversion from the QC to QCO dialect here.

I took the liberty to remove the QC builder and the conversion. Make sure to study the changes!

copy that!

this might lead to easier comparison of original and altered circuit as well. i can compare the indices of the static qubits afterwards as they should be changed by the pass, am i correct with this assumption? currently working on that to setup more test cases ...

I think the easiest way to check this - without having to worry about the order of the static operations¹ - is to verify that the quantum operations use the correct static indices as input.

Example:

0 ---X---[H]---.---<sink> | | 1 ---X---------⨁---<sink> ↓ (absorb SWAPs) 1 ---[H]---.---<sink> // Input of `HOp` is `qco.static 1`? | // 0 ---------⨁---<sink> // Input of `CtrlOp(target)` is `qco.static 0`?

Footnotes

Note that the order (from top to bottom) of the qco.static operations doesn't matter. ↩

works just fine 👍

i have the changes in my fork at the moment, is there any possibility to append the commit to this very PR somehow or do i have to create a separate PR from my fork?

Although suboptimal, I locally merged your changes into the branch of the main repository. Seems to work.

Maybe once you think you are done, ping me again here in this thread and I will re-merge the changes 👀

got it, fine with me :)

MatthiasReumann · 2026-05-04T07:21:31Z

+  }
+
+private:
+  static void insertStatics(ModuleOp anchor, IRRewriter& rewriter) {


This function isn't required anymore. Instead of using allocations in the unit tests simply use static qubits (via qco.static).

MatthiasReumann · 2026-05-04T07:23:09Z


+def SwapAbsorptionPass : Pass<"absorb-swaps", "mlir::ModuleOp"> {
+  let dependentDialects = ["mlir::qco::QCODialect"];
+  let summary = "";


Make sure to add a nice summary and description of the swap-absorption pass here.

…-quantum-toolkit/core into feat/swap-absorption-pass

github-actions · 2026-05-06T05:43:36Z

Cpp-Linter Report ⚠️

Some files did not pass the configured checks!

clang-tidy (v22.1.3) reports: 4 concern(s)

mlir/lib/Dialect/QCO/Transforms/Optimizations/SwapAbsorption.cpp:15:1: warning: [misc-include-cleaner]

included header QTensorOps.h is not used directly
```
   15 | #include "mlir/Dialect/QTensor/IR/QTensorOps.h"
      | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   16 | 
```
mlir/lib/Dialect/QCO/Transforms/Optimizations/SwapAbsorption.cpp:17:1: warning: [misc-include-cleaner]

included header STLExtras.h is not used directly
```
   17 | #include <llvm/ADT/STLExtras.h>
      | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   18 | #include <mlir/Dialect/Func/IR/FuncOps.h>
```
mlir/lib/Dialect/QCO/Transforms/Optimizations/SwapAbsorption.cpp:21:1: warning: [misc-include-cleaner]

included header Visitors.h is not used directly
```
   21 | #include <mlir/IR/Visitors.h>
      | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   22 | #include <mlir/Support/LLVM.h>
```
mlir/lib/Dialect/QCO/Transforms/Optimizations/SwapAbsorption.cpp:25:1: warning: [misc-include-cleaner]

included header cstddef is not used directly
```
   25 | #include <cstddef>
      | ^~~~~~~~~~~~~~~~~~
   26 | #include <utility>
```

Have any feedback or feature suggestions? Share it here.

jmoosburger · 2026-05-06T20:20:05Z

a topic for another meeting might be all the checks and pipelines in the PRs. the Lint stage fails currently even if the pre-commit fixes fixed some linting stuff. (the Check failed because of the Lint-step i guess?)

locally i use clang-tidy now

burgholzer · 2026-05-07T07:02:50Z

a topic for another meeting might be all the checks and pipelines in the PRs. the Lint stage fails currently even if the pre-commit fixes fixed some linting stuff. (the Check failed because of the Lint-step i guess?)

locally i use clang-tidy now

You can always see the reason for lint job failures in the CI summary: https://github.com/munich-quantum-toolkit/core/actions/runs/25418523584/attempts/1#summary-74555188566
In this case, there are 4 open clang-tidy warnings.

Add SWAPAbsorption Pass

00ad822

MatthiasReumann self-assigned this May 4, 2026

🎨 pre-commit fixes

77250d1

MatthiasReumann mentioned this pull request May 4, 2026

✨ Add -absorb-swaps pass #1661

Closed

11 tasks

MatthiasReumann commented May 4, 2026

View reviewed changes

MatthiasReumann added c++ Anything related to C++ code MLIR Anything related to MLIR labels May 4, 2026

MatthiasReumann and others added 7 commits May 4, 2026 09:32

Fix linting

515e6bf

Merge branch 'feat/swap-absorption-pass' of https://github.com/munich…

f6e1c39

…-quantum-toolkit/core into feat/swap-absorption-pass

🎨 pre-commit fixes

b0474ff

remove SwapAbsorbtion::insertStatics()

10302c5

change unittests to use static qubits

da710c7

SwapAbsorbtion: PassReordersTwoQubitCircuitWithLeadingSwap

d793c73

🎨 pre-commit fixes

de89411

Uh oh!

Conversation

MatthiasReumann commented May 4, 2026

Description

Checklist

Uh oh!

codecov Bot commented May 4, 2026

Codecov Report

Uh oh!

MatthiasReumann left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Footnotes

Uh oh!

jmoosburger May 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

github-actions Bot commented May 6, 2026

Cpp-Linter Report ⚠️

Uh oh!

jmoosburger commented May 6, 2026

Uh oh!

burgholzer commented May 7, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

MatthiasReumann left a comment •

edited

Loading

jmoosburger May 5, 2026 •

edited

Loading