Process V3 save branch on restore to advance PC past branch data

The Quetzal spec stores the PC pointing at the save instruction's
branch data. On restore, this branch must be processed as "save
succeeded" to advance the PC to the actual next instruction. Without
this, the branch bytes were decoded as an opcode, corrupting execution.

Detect the save opcode (0xB5) immediately before the restored PC to
distinguish in-game saves from out-of-band saves (which don't need
branch processing). Also improve error diagnostics: pop_stack now
raises ZStackPopError with frame context, and the instruction trace
dumps on all exceptions.

src/mudlib/embedded_if_session.py

@@ -53,6 +53,17 @@ class EmbeddedIFSession:
             save_data = self.save_path.read_bytes()
             parser = QuetzalParser(self._zmachine)
             parser.load_from_bytes(save_data)
+            # In V1-3, the saved PC points to branch data after the save
+            # instruction. Process the branch as "save succeeded" so the
+            # PC advances past it. Detect by checking for save opcode (0xB5)
+            # immediately before the restored PC.
+            pc = self._zmachine._opdecoder.program_counter
+            if (
+                self._zmachine._mem.version <= 3
+                and pc > 0
+                and self._zmachine._mem[pc - 1] == 0xB5
+            ):
+                self._zmachine._cpu._branch(True)
             return True
         except Exception as e:
             logger.debug(f"Restore failed: {e}")

src/mudlib/zmachine/zcpu.py

@@ -236,8 +236,7 @@ class ZCpu:
         except (ZCpuQuit, ZCpuRestart):
             # Normal control flow - don't dump trace
             raise
-        except ZCpuError:
+        except Exception:
-            # All other ZCpu errors - dump trace for debugging
             self._dump_trace()
             raise
         return True

src/mudlib/zmachine/zstackmanager.py

@@ -153,6 +153,11 @@ class ZStackManager:
         "Remove and return value from the top of the data stack."
 
         current_routine = self._call_stack[-1]
+        if not current_routine.stack:
+            frame_idx = len(self._call_stack) - 1
+            ra = getattr(current_routine, "return_addr", "N/A")
+            pc = getattr(current_routine, "program_counter", 0)
+            raise ZStackPopError(f"frame {frame_idx}, return_addr={ra}, pc=0x{pc:06x}")
         return current_routine.stack.pop()
 
     def get_stack_frame_index(self):