[c++] Mid-circuit measurements inside c-like functions generate incorrect results

Required prerequisites

• Consult the security policy. If reporting a security vulnerability, do not report the bug using this form. Use the process described in the policy to report the issue.
• Make sure you've read the documentation. Your issue may be addressed there.
• 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

Mid-circuit measurements inside c-like functions generate incorrect results due to cudaq::getKernelName returning incorrect name ("void ()")

Related: #3160

Steps to reproduce the bug

/*******************************************************************************
 * Copyright (c) 2022 - 2025 NVIDIA Corporation & Affiliates.                  *
 * All rights reserved.                                                        *
 *                                                                             *
 * This source code and the accompanying materials are made available under    *
 * the terms of the Apache License 2.0 which accompanies this distribution.    *
 ******************************************************************************/

// clang-format off
// Simulators
// RUN: nvq++ %cpp_std --enable-mlir  %s -o %t && %t | FileCheck %s
// RUN: nvq++ %cpp_std --library-mode %s -o %t && %t | FileCheck %s

// Quantum emulators
// RUN: nvq++ %cpp_std --target quantinuum               --emulate %s -o %t && %t | FileCheck %s
// RUN: nvq++ %cpp_std --target ionq                     --emulate %s -o %t && %t | FileCheck %s
// RUN: nvq++ %cpp_std --target anyon                    --emulate %s -o %t && %t | FileCheck %s
// 2 different IQM machines for 2 different topologies
// RUN: nvq++ %cpp_std --target iqm --iqm-machine Adonis --emulate %s -o %t && %t | FileCheck %s
// RUN: nvq++ %cpp_std --target iqm --iqm-machine Apollo --emulate %s -o %t && %t | FileCheck %s
// RUN: nvq++ %cpp_std --target oqc                      --emulate %s -o %t && %t | FileCheck %s
// RUN: if %braket_avail; then nvq++ %cpp_std --target braket --emulate %s -o %t && %t | FileCheck %s; fi
// clang-format on

#include <cudaq.h>
#include <iostream>
#include <string>
#include <vector>

__qpu__ void callee(cudaq::qview<> r)  {
  for (auto i = 0; i < 4; i++) {
    if (i % 2 == 0)
        x(r[i]);

    auto m = mz(r[i]);
    cudaq::reset(r[i]);

    if (m)
        x(r[i]);
    else
        h(r[i]);
  }
}

__qpu__ void c_caller()  {
  cudaq::qvector q(4);
    callee(q);
}

__qpu__ void c_inlined() {
  cudaq::qvector r(4);
  for (auto i = 0; i < 4; i++) {
    if (i % 2 == 0)
        x(r[i]);

    auto m = mz(r[i]);
    cudaq::reset(r[i]);

    if (m)
        x(r[i]);
    else
        h(r[i]);
  }
}


int main() {
  {
    auto counts = cudaq::sample(1000, c_caller);
    counts.dump();

    printf("%zu\n", counts.count("1010", "__global__"));
    printf("%zu\n", counts.count("1011", "__global__"));
    printf("%zu\n", counts.count("1110", "__global__"));
    printf("%zu\n", counts.count("1111", "__global__"));
  }

  {
    auto counts = cudaq::sample(1000, c_inlined);
    counts.dump();

    printf("%zu\n", counts.count("1010", "__global__"));
    printf("%zu\n", counts.count("1011", "__global__"));
    printf("%zu\n", counts.count("1110", "__global__"));
    printf("%zu\n", counts.count("1111", "__global__"));
  }
  return 0;
}

Output:
0
0
0
1000

Expected behavior

Output:

255
255
255
255

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

Not a regression

Environment

• CUDA-Q version:
• Python version:
• C++ compiler:
• Operating system:

Suggestions

No response

Start with the supplied C++ reproduction, especially the c_caller and c_inlined entry points, and run its listed nvq++ and FileCheck commands. Trace cudaq::getKernelName where it returns "void ()"; done means the reproduced counts match the expected 255 values for all four bitstrings.