✨ Implement two-qubit gate decomposition pass#1206

Draft

taminob wants to merge 163 commits intomainfrom

taminob/mlir-two-qubit-gate-decomposition-pass

Commits on Dec 28, 2025

snapshot
taminob
committed
copied rust code
taminob
committed
🎨 pre-commit fixes

pre-commit-ci[bot]
authored and
taminob
committed
snapshot
taminob
committed
snapshot
taminob
committed
snapshot
taminob
committed
minor additions
taminob
committed
add pass to Passes.td
taminob
committed
snapshot
taminob
committed
compiling snapshot
taminob
committed
tmp
taminob
committed
switch to long double
taminob
committed
testing a.cpp
taminob
committed
add zgemm for testing, start euler decomposition implementation
taminob
committed
non-working but finished snapshot
taminob
committed
fix out-of-bounds issues
taminob
committed
test using Armadillo
taminob
committed
minor fix
taminob
committed
tmp
taminob
committed
work on gate application
taminob
committed
use laughing-umbrella implementation for decompose_two_qubit_product_gate
taminob
committed
add more eigen decomposition implementations
taminob
committed
work on applySeries()
taminob
committed
Revert "use laughing-umbrella implementation for decompose_two_qubit_product_gate"
taminob
committed
fix transpose_conjugate
taminob
committed
add test
taminob
committed
actually kind of working snapshot?
taminob
committed
fix test case
taminob
committed
minor fix
taminob
committed
fix test case
taminob
committed
add new test, fix tests
taminob
committed
iniital complexSeries test
taminob
committed
use Eigen for multiply() and determinant()
taminob
committed
fix gate matrix multiplication
taminob
committed
add simple (too simple) series complexity
taminob
committed
fix insertion when operations which are not part of the series are involved
taminob
committed
add gate matrices
taminob
committed
sequence complexity()
taminob
committed
print debugging :)
taminob
committed
longer complex test to pass sequence complexity check
taminob
committed
use CXGate() definition for decomposer
taminob
committed
add cnot test other direction
taminob
committed
tmp (to be deleted)
taminob
committed
re-write using Eigen
taminob
committed
delete pulse-optimizer code (KEEP for future; should be re-added at some point)
taminob
committed
use designated initializer lists
taminob
committed
fix warnings, rename all variables
taminob
committed
test normalized magic basis, allow complex eigenvector decomposition
taminob
committed
finally figured some stuff out
taminob
committed
try out lauging-umbrella weyl-chamber
taminob
committed
try to implement weylchamber.c1c2c3 algorithm
taminob
committed
on_gates two-qubit product decompose
taminob
committed
minor additions
taminob
committed
try ordering all four eigenvectors
taminob
committed
undo ordering of all columns
taminob
committed
implement algorithm based on quantumflow.canonical_decomposition
taminob
committed
sanity checks and stuff
taminob
committed
add original implementation again which seems to kind of work?
taminob
committed
generate openqasm code of generated sequence
taminob
committed
some more small fixes, maybe breakthrough?
taminob
committed
rework getTwoQubitSeries logic
taminob
committed
generate openqasm code of found series
taminob
committed
fix new getTwoQubitSeries logic
taminob
committed
working except global phase?
taminob
committed
fix global phase issues
taminob
committed
first working snapshot
taminob
committed
very minor cleanup
taminob
committed
add more sanity checks for unitary
taminob
committed
find and fix minor formula mistake
taminob
committed
test update
taminob
committed
make calculations exact without approximation
taminob
committed
respect gphase of circuit and fix series collection bug
taminob
committed
change global phase to local phase
taminob
committed
revert to global phase, fix controlled gate qubit order
taminob
committed
fix first couple of tests
taminob
committed
start working on basis set and remove dead code
taminob
committed
decompose for all bases, choose best
taminob
committed
add checks for most tests
taminob
committed
TMP, needs to be removed; allow OpenQASM export with round-trip pass
taminob
committed
add single-qubit backtracking
taminob
committed
fix atol issues
taminob
committed
add more euler bases to pattern
taminob
committed
minor cleanup (renaming, comments, disable printing)
taminob
committed
Revert "TMP, needs to be removed; allow OpenQASM export with round-trip pass"
taminob
committed
remove armadillo in cmake
taminob
committed
more cleanup
taminob
committed
fix (?) tests
taminob
committed
fix final test
taminob
committed
another minor cleanup
taminob
committed
clean sanity check prints up
taminob
committed
properly add Eigen lib to cmake
taminob
committed
remove unnecessary correction
taminob
committed
🎨 pre-commit fixes

pre-commit-ci[bot]
authored and
taminob
committed
fix eigen compilation issues
taminob
committed
more cleanup
taminob
committed
clean up weyl decomposition creation
taminob
committed
remove openqasm code generation
taminob
committed
fix includes and use SmallVector instead of std::vector
taminob
committed
fix or silence linter
taminob
committed
incorporate coderabbitai feedback
taminob
committed
remove global phase accumulation
taminob
committed
fix :math: latex formatting
taminob
committed
minor comment fixes
taminob
committed
replace CHECK-DAG by CHECK again
taminob
committed
add comments to weyl decomposition
taminob
committed
extract TwoQubitWeylDecomposition::getTrace()
taminob
committed
transform basis decomposer into actual class with private members
taminob
committed
attempt to fix CI test
taminob
committed
add supercontrolled check
taminob
committed
extract decomposition into headers
taminob
committed
slight improvement to weyl decomposition for testing purposes
taminob
committed
add unittest for weyl decomposition
taminob
committed
fix linter issues
taminob
committed
add tests for basis decomposer
taminob
committed
fix linter issues
taminob
committed
clang-tidy: Add IgnoreHeader for include cleaner for Eigen
taminob
committed
remove exact values from MLIR test
taminob
committed
fix mlir test/linter issues
taminob
committed
remove std::numbers because not supported by all compilers in CI
taminob
committed
remove unused Helpers.h from src dir
taminob
committed
remove auto* because of windows build
taminob
committed
try to fix windows ARM ci
taminob
committed
pass matrices by reference
taminob
committed
allow non-optimal vectorization
taminob
committed
🎨 pre-commit fixes

pre-commit-ci[bot]
authored and
taminob
committed
try to fix
taminob
committed
integrate single-qubit decomposition
taminob
committed
add tests for single-qubit decomposition
taminob
committed
add unittests for euler decomposition
taminob
committed
fix linter issues
taminob
committed
fix include cleaner issues
taminob
committed
add test parametrizations for other specializations
taminob
committed
add more matrix definitions
taminob
committed
handle barrier gates in series collection
taminob
committed
🎨 pre-commit fixes

pre-commit-ci[bot]
authored and
taminob
committed
fix include cleaner issues
taminob
committed
remove unused globalPhase from TwoQubitSeries
taminob
committed
fix barrier handling at start of series
taminob
committed
simplify weyl specialization test parameters
taminob
committed
remove dynamic_casts due to RTTI issue
taminob
committed
Revert "remove dynamic_casts due to RTTI issue"
taminob
committed
move GateDecompositionPass to new QCO dialect
taminob
committed
add GateDecompositionPass to compiler pipeline
taminob
committed
🎨 pre-commit fixes

pre-commit-ci[bot]
authored and
taminob
committed
add UnitaryMatrixType to operation and add Eigen dependency to dialect
taminob
committed
first non-working attempt
taminob
committed

Commits on Jan 4, 2026

second non-working attempt
taminob
committed
compiling approach
taminob
committed

Uh oh!

Commits

Commits on Dec 28, 2025