Allow for more complex builder gate signatures in library mode

Required prerequisites

[X] Make sure you've read the documentation. Your issue may be addressed there.
[X] Search the issue tracker to verify that this hasn't already been reported. +1 or comment there if it has.
[ ] If possible, make a PR with a failing test to give us a starting point to work on!

Describe the bug

The following code only passes in emulation, but not in library mode. The bug is a qpp dimension issue (more triage needed to determine what impact, if any, on nvidia backends), meaning we will likely need to update somewhere along the function stack of qpp::applyGate.

#include "cudaq.h"
#include <iostream>

// fails with: `nvq++ test.cpp && ./a.out`
// passes with: `nvq++ --target quantinuum --emulate test.cpp && ./a.out`

int main() {
    auto kernel = cudaq::make_kernel();
    // Using a mixture of qregs and qubits as controls
    auto controls1 = kernel.qalloc(2);
    auto controls2 = kernel.qalloc(2);
    auto control3 = kernel.qalloc();
    auto target = kernel.qalloc();

    // bug with any control gate
    kernel.x<cudaq::ctrl>(controls1, controls2, control3, target);

    auto counts = cudaq::sample(kernel);
    counts.dump();
}

Runtime error:

terminate called after throwing an instance of 'qpp::exception::SubsysMismatchDims'
  what():  qpp::applyCTRL(): Subsystems mismatch dimensions! [ctrl/dims]

Steps to reproduce the bug

Run the above code with nvq++ test.cpp && ./a.out

Expected behavior

The behavior should match the output of nvq++ --target quantinuum --emulate test.cpp && ./a.out .

Note: the reason this is an issue only in library mode is because of the compiler passes that are specific to emulation. They will redefine the multi-controls as two qubit gates, circumventing the dimension issue.

Is this a regression? If it is, put the last known working version (or commit) here.

Not a regression

Environment

CUDA Quantum version: latest
Python version: latest
C++ compiler: lates
Operating system: Ubuntu

Suggestions

No response

NVIDIA / cuda-quantum