Trace a Function Call Through CPython

Tracing a Python Call

A Python function call crosses several C layers before your bytecode runs. Understanding \`PyObject_Call\`, frame setup, and the eval loop explains stack traces, profiling, and debugger breakpoints.

The Call Chain

Rough path from \`f()\` to execution:

1. \`PyObject_Call\` or \`PyObject_Vectorcall\` (fast path for simple calls)
2. Allocate a \`PyFrameObject\` with locals, stack, and code object
3. Enter \`_PyEval_EvalFrameDefault\` (the ceval loop)
4. On return, unwind the frame and propagate the result

Each frame holds:

• Reference to the \`PyCodeObject\` (bytecode, constants, names)
• Operand stack for the VM
• \`f_locals\` and fast locals array

For example, calling \`add(2, 3)\` pushes arguments onto the new frame's stack, runs \`LOAD_FAST\`/\`BINARY_ADD\`, then returns via \`RETURN_VALUE\`.

Design Notes

Frame objects are among the hottest allocations in CPython. The interpreter caches empty frames and reuses them where possible. Watch `PyFrame_New` in profiles when optimizing call-heavy Python code.

For This Exercise

This exercise asks you to document the CPython call path from Python source to \`ceval.c\`. You will trace frame creation, argument binding, and return so you can read real interpreter source with context.

You will use the same mental model here when reading production interpreter source later in the track. Sketch one concrete input on paper, predict the outcome, then confirm with code. That discipline catches logic errors early and makes debugging far faster when you extend the implementation in follow-on tasks.