GenericBackendV2 should fail when the backend cannot allocate the bas…

…is gate because its size (backport #12653) (#12666) * GenericBackendV2 should fail when the backend cannot allocate the basis gate because its size (#12653) * GenericBackendV2 should fail when the backend cannot allocate the basis gate because its size Co-authored-by: Elena Peña Tapia <[email protected]> * reno * Update releasenotes/notes/fixes_GenericBackendV2-668e40596e1f070d.yaml Co-authored-by: Elena Peña Tapia <[email protected]> * another single qubit backend --------- Co-authored-by: Elena Peña Tapia <[email protected]> (cherry picked from commit e36027c) # Conflicts: # qiskit/providers/fake_provider/generic_backend_v2.py # test/visual/mpl/graph/test_graph_matplotlib_drawer.py * Fix Conflict * Update test_graph_matplotlib_drawer.py * black --------- Co-authored-by: Luciano Bello <[email protected]> Co-authored-by: Elena Peña Tapia <[email protected]>
Qiskit · Jun 26, 2024 · 5def925 · 5def925
1 parent 6cc5796
commit 5def925
Show file tree

Hide file tree

Showing 3 changed files with 20 additions and 0 deletions.
diff --git a/qiskit/providers/fake_provider/generic_backend_v2.py b/qiskit/providers/fake_provider/generic_backend_v2.py
@@ -368,6 +368,12 @@ def _build_generic_target(self):
                     f"in the standard qiskit circuit library."
                 )
             gate = self._supported_gates[name]
+
+            if self.num_qubits < gate.num_qubits:
+                raise QiskitError(
+                    f"Provided basis gate {name} needs more qubits than {self.num_qubits}, "
+                    f"which is the size of the backend."
+                )
             noise_params = self._get_noise_defaults(name)
             self._add_noisy_instruction_to_target(gate, noise_params, calibration_inst_map)
 

diff --git a/releasenotes/notes/fixes_GenericBackendV2-668e40596e1f070d.yaml b/releasenotes/notes/fixes_GenericBackendV2-668e40596e1f070d.yaml
@@ -0,0 +1,4 @@
+---
+fixes:
+  - |
+    The constructor :class:`.GenericBackendV2` was allowing to create malformed backends because it accepted basis gates that couldn't be allocated in the backend size . That is, a backend with a single qubit should not accept a basis with two-qubit gates.
diff --git a/test/python/providers/fake_provider/test_generic_backend_v2.py b/test/python/providers/fake_provider/test_generic_backend_v2.py
@@ -36,6 +36,16 @@ def test_supported_basis_gates(self):
         with self.assertRaises(QiskitError):
             GenericBackendV2(num_qubits=8, basis_gates=["cx", "id", "rz", "sx", "zz"])
 
+    def test_cx_1Q(self):
+        """Test failing with a backend with single qubit but with a two-qubit basis gate"""
+        with self.assertRaises(QiskitError):
+            GenericBackendV2(num_qubits=1, basis_gates=["cx", "id"])
+
+    def test_ccx_2Q(self):
+        """Test failing with a backend with two qubits but with a three-qubit basis gate"""
+        with self.assertRaises(QiskitError):
+            GenericBackendV2(num_qubits=2, basis_gates=["ccx", "id"])
+
     def test_operation_names(self):
         """Test that target basis gates include "delay", "measure" and "reset" even
         if not provided by user."""