Abstract: A dual-mode debugger can debug native code or interpreted code. Transitioning between modes can be triggered by defined bailout conditions. A program can be debugged by attaching a debugger and compiling the program into native code under the debugger. At defined points in the program, debugging can transition from native mode to interpreted mode. Debugging the native code can stop, current state can be captured, and an interpreter can be instantiated. An interpreter stack can be created and can be populated with the captured current state. Debugging operations which involve execution control: pausing execution, continuing execution, stepping into, out of or over a section of code and so on, can occur in interpreted mode, which is typically easier to implement. Debugging operations which involve inspection and modification of variables can occur in native mode.

Abstract: A dual-mode debugger can debug native code or interpreted code. Transitioning between modes can be triggered by defined bailout conditions. A program can be debugged by attaching a debugger and compiling the program into native code under the debugger. At defined points in the program, debugging can transition from native mode to interpreted mode. Debugging the native code can stop, current state can be captured, and an interpreter can be instantiated. An interpreter stack can be created and can be populated with the captured current state. Debugging operations which involve execution control: pausing execution, continuing execution, stepping into, out of or over a section of code and so on, can occur in interpreted mode, which is typically easier to implement. Debugging operations which involve inspection and modification of variables can occur in native mode.