Add support for scf.for ops in the convert-quantum-to-qecl pass

[joeycarter]

Context: In order to support more general input programs in the experimental QEC compilation pipeline, we must add support for control-flow operations.

Description of the Change: This PR adds support for for loops (via scf.for ops) in the convert-quantum-to-qecl pass. For example, an input program such as

func.func @test_program() {
    %c1 = arith.constant 1 : index
    %c5 = arith.constant 5 : index
    %c0 = arith.constant 0 : index
    %0 = quantum.alloc(1) : !quantum.reg
    %1 = quantum.extract %0[0] : !quantum.reg -> !quantum.bit
    %2 = quantum.custom "PauliX"() %1 : !quantum.bit
    %3 = scf.for %arg0 = %c0 to %c5 step %c1 iter_args(%arg1 = %2) -> (!quantum.bit) {
        %out_qubits = quantum.custom "PauliX"() %arg1 : !quantum.bit
        scf.yield %out_qubits : !quantum.bit
    }
    %4 = quantum.insert %0[0], %3 : !quantum.reg, !quantum.bit
    quantum.dealloc %4 : !quantum.reg
    func.return
}

is lowered to

func.func @test_program() {
    %c1 = arith.constant 1 : index
    %c5 = arith.constant 5 : index
    %c0 = arith.constant 0 : index
    %0 = qecl.alloc() : !qecl.hyperreg<1 x 1>
    %1 = qecl.extract_block %0[0] : !qecl.hyperreg<1 x 1> -> !qecl.codeblock<1>
    %2 = qecl.encode[zero] %1 : !qecl.codeblock<1>
    %3 = qecl.insert_block %0[0], %2 : !qecl.hyperreg<1 x 1>, !qecl.codeblock<1>
    %4 = qecl.extract_block %3[0] : !qecl.hyperreg<1 x 1> -> !qecl.codeblock<1>
    %5 = qecl.qec %4 : !qecl.codeblock<1>
    %6 = qecl.x %5[0] : !qecl.codeblock<1>
    %7 = qecl.qec %6 : !qecl.codeblock<1>
    %8 = scf.for %arg0 = %c0 to %c5 step %c1 iter_args(%arg1 = %7) -> (!qecl.codeblock<1>) {
        %out_qubits = qecl.x %arg1[0] : !qecl.codeblock<1>
        %out_qubits_1 = qecl.qec %out_qubits : !qecl.codeblock<1>
        scf.yield %out_qubits_1 : !qecl.codeblock<1>
    }
    %9 = qecl.insert_block %3[0], %8 : !qecl.hyperreg<1 x 1>, !qecl.codeblock<1>
    qecl.dealloc %9 : !qecl.hyperreg<1 x 1>
    func.return
}

Support for scf.for ops that return !quantum.reg values, and a mix of quantum and classical types, has also been added.

[sc-120266]

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.04%. Comparing base (ab54da1) to head (89a9673).

Additional details and impacted files

@@                        Coverage Diff                         @@
##           joeycarter/quantum-to-qecl/scf-if    #2881   +/-   ##
==================================================================
  Coverage                              97.03%   97.04%           
==================================================================
  Files                                    167      167           
  Lines                                  18876    18914   +38     
  Branches                                1766     1767    +1     
==================================================================
+ Hits                                   18317    18355   +38     
  Misses                                   415      415           
  Partials                                 144      144           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[multiphaseCFD]

looks like the current implementation is bounded to k=1? Any implementation challenges you can foresee if we move to k > 1.

[joeycarter]

In fact, I think the control-flow conversion, as implemented here, should be generalizable to k > 1. The piece that would need to be updated is the conversion pattern for the quantum.extract op, which currently assumes k = 1. The tests still use k = 1 because the pass itself still raises an error when given k != 1 🙂

[joeycarter]

But of course we'd need to verify once we start working on support for k > 1!

[multiphaseCFD]

Should be lots of fun for k>1 , would like to know more about that later. Feel free to tag me to take a look