2024-06-26 09:18:20 -03:00
|
|
|
# Frames
|
2022-04-11 12:05:20 -03:00
|
|
|
|
2024-06-26 09:18:20 -03:00
|
|
|
Each call to a Python function has an activation record, commonly known as a
|
|
|
|
"frame". It contains information about the function being executed, consisting
|
|
|
|
of three conceptual sections:
|
2022-04-11 12:05:20 -03:00
|
|
|
|
|
|
|
* Local variables (including arguments, cells and free variables)
|
|
|
|
* Evaluation stack
|
2024-06-26 09:18:20 -03:00
|
|
|
* Specials: The per-frame object references needed by the VM, including
|
|
|
|
globals dict, code object, instruction pointer, stack depth, the
|
|
|
|
previous frame, etc.
|
|
|
|
|
2024-10-21 19:37:31 -03:00
|
|
|
The definition of the `_PyInterpreterFrame` struct is in
|
|
|
|
[Include/internal/pycore_frame.h](../Include/internal/pycore_frame.h).
|
2022-04-11 12:05:20 -03:00
|
|
|
|
2024-06-26 09:18:20 -03:00
|
|
|
# Allocation
|
2022-04-11 12:05:20 -03:00
|
|
|
|
2024-06-26 09:18:20 -03:00
|
|
|
Python semantics allows frames to outlive the activation, so they need to
|
|
|
|
be allocated outside the C call stack. To reduce overhead and improve locality
|
|
|
|
of reference, most frames are allocated contiguously in a per-thread stack
|
2024-10-21 19:37:31 -03:00
|
|
|
(see `_PyThreadState_PushFrame` in [Python/pystate.c](../Python/pystate.c)).
|
2024-06-26 09:18:20 -03:00
|
|
|
|
|
|
|
Frames of generators and coroutines are embedded in the generator and coroutine
|
2024-10-21 19:37:31 -03:00
|
|
|
objects, so are not allocated in the per-thread stack. See `PyGenObject` in
|
|
|
|
[Include/internal/pycore_genobject.h](../Include/internal/pycore_genobject.h).
|
2024-06-26 09:18:20 -03:00
|
|
|
|
|
|
|
## Layout
|
2022-04-11 12:05:20 -03:00
|
|
|
|
|
|
|
Each activation record is laid out as:
|
2024-06-26 09:18:20 -03:00
|
|
|
* Specials
|
2022-04-11 12:05:20 -03:00
|
|
|
* Locals
|
|
|
|
* Stack
|
|
|
|
|
|
|
|
This seems to provide the best performance without excessive complexity.
|
2024-06-26 09:18:20 -03:00
|
|
|
The specials have a fixed size, so the offset of the locals is know. The
|
|
|
|
interpreter needs to hold two pointers, a frame pointer and a stack pointer.
|
2022-04-11 12:05:20 -03:00
|
|
|
|
|
|
|
#### Alternative layout
|
|
|
|
|
|
|
|
An alternative layout that was used for part of 3.11 alpha was:
|
|
|
|
|
|
|
|
* Locals
|
2024-06-26 09:18:20 -03:00
|
|
|
* Specials
|
2022-04-11 12:05:20 -03:00
|
|
|
* Stack
|
|
|
|
|
|
|
|
This has the advantage that no copying is required when making a call,
|
|
|
|
as the arguments on the stack are (usually) already in the correct
|
|
|
|
location for the parameters. However, it requires the VM to maintain
|
|
|
|
an extra pointer for the locals, which can hurt performance.
|
|
|
|
|
|
|
|
### Generators and Coroutines
|
|
|
|
|
|
|
|
Generators and coroutines contain a `_PyInterpreterFrame`
|
|
|
|
The specials sections contains the following pointers:
|
|
|
|
|
|
|
|
* Globals dict
|
|
|
|
* Builtins dict
|
|
|
|
* Locals dict (not the "fast" locals, but the locals for eval and class creation)
|
|
|
|
* Code object
|
|
|
|
* Heap allocated `PyFrameObject` for this activation record, if any.
|
|
|
|
* The function.
|
|
|
|
|
|
|
|
The pointer to the function is not strictly required, but it is cheaper to
|
|
|
|
store a strong reference to the function and borrowed references to the globals
|
|
|
|
and builtins, than strong references to both globals and builtins.
|
|
|
|
|
|
|
|
### Frame objects
|
|
|
|
|
|
|
|
When creating a backtrace or when calling `sys._getframe()` the frame becomes
|
2022-11-10 08:34:57 -04:00
|
|
|
visible to Python code. When this happens a new `PyFrameObject` is created
|
2022-04-11 12:05:20 -03:00
|
|
|
and a strong reference to it placed in the `frame_obj` field of the specials
|
|
|
|
section. The `frame_obj` field is initially `NULL`.
|
|
|
|
|
|
|
|
The `PyFrameObject` may outlive a stack-allocated `_PyInterpreterFrame`.
|
|
|
|
If it does then `_PyInterpreterFrame` is copied into the `PyFrameObject`,
|
|
|
|
except the evaluation stack which must be empty at this point.
|
2024-06-26 09:18:20 -03:00
|
|
|
The previous frame link is updated to reflect the new location of the frame.
|
2022-04-11 12:05:20 -03:00
|
|
|
|
|
|
|
This mechanism provides the appearance of persistent, heap-allocated
|
|
|
|
frames for each activation, but with low runtime overhead.
|
|
|
|
|
|
|
|
### Generators and Coroutines
|
|
|
|
|
2024-10-21 19:37:31 -03:00
|
|
|
Generators (objects of type `PyGen_Type`, `PyCoro_Type` or
|
|
|
|
`PyAsyncGen_Type`) have a `_PyInterpreterFrame` embedded in them, so
|
2024-06-26 09:18:20 -03:00
|
|
|
that they can be created with a single memory allocation.
|
|
|
|
When such an embedded frame is iterated or awaited, it can be linked with
|
|
|
|
frames on the per-thread stack via the linkage fields.
|
2022-04-11 12:05:20 -03:00
|
|
|
|
|
|
|
If a frame object associated with a generator outlives the generator, then
|
2024-06-26 09:18:20 -03:00
|
|
|
the embedded `_PyInterpreterFrame` is copied into the frame object (see
|
2024-10-21 19:37:31 -03:00
|
|
|
`take_ownership()` in [Python/frame.c](../Python/frame.c)).
|
2022-04-11 12:05:20 -03:00
|
|
|
|
|
|
|
### Field names
|
|
|
|
|
|
|
|
Many of the fields in `_PyInterpreterFrame` were copied from the 3.10 `PyFrameObject`.
|
|
|
|
Thus, some of the field names may be a bit misleading.
|
|
|
|
|
|
|
|
For example the `f_globals` field has a `f_` prefix implying it belongs to the
|
|
|
|
`PyFrameObject` struct, although it belongs to the `_PyInterpreterFrame` struct.
|
2024-06-26 09:18:20 -03:00
|
|
|
We may rationalize this naming scheme for a later version.
|
2022-11-10 08:34:57 -04:00
|
|
|
|
|
|
|
|
|
|
|
### Shim frames
|
|
|
|
|
|
|
|
On entry to `_PyEval_EvalFrameDefault()` a shim `_PyInterpreterFrame` is pushed.
|
|
|
|
This frame is stored on the C stack, and popped when `_PyEval_EvalFrameDefault()`
|
|
|
|
returns. This extra frame is inserted so that `RETURN_VALUE`, `YIELD_VALUE`, and
|
|
|
|
`RETURN_GENERATOR` do not need to check whether the current frame is the entry frame.
|
|
|
|
The shim frame points to a special code object containing the `INTERPRETER_EXIT`
|
|
|
|
instruction which cleans up the shim frame and returns.
|
2023-10-26 10:43:10 -03:00
|
|
|
|
|
|
|
|
|
|
|
### The Instruction Pointer
|
|
|
|
|
|
|
|
`_PyInterpreterFrame` has two fields which are used to maintain the instruction
|
|
|
|
pointer: `instr_ptr` and `return_offset`.
|
|
|
|
|
|
|
|
When a frame is executing, `instr_ptr` points to the instruction currently being
|
|
|
|
executed. In a suspended frame, it points to the instruction that would execute
|
|
|
|
if the frame were to resume. After `frame.f_lineno` is set, `instr_ptr` points to
|
|
|
|
the next instruction to be executed. During a call to a python function,
|
|
|
|
`instr_ptr` points to the call instruction, because this is what we would expect
|
|
|
|
to see in an exception traceback.
|
|
|
|
|
|
|
|
The `return_offset` field determines where a `RETURN` should go in the caller,
|
|
|
|
relative to `instr_ptr`. It is only meaningful to the callee, so it needs to
|
|
|
|
be set in any instruction that implements a call (to a Python function),
|
|
|
|
including CALL, SEND and BINARY_SUBSCR_GETITEM, among others. If there is no
|
|
|
|
callee, then return_offset is meaningless. It is necessary to have a separate
|
|
|
|
field for the return offset because (1) if we apply this offset to `instr_ptr`
|
|
|
|
while executing the `RETURN`, this is too early and would lose us information
|
|
|
|
about the previous instruction which we could need for introspecting and
|
|
|
|
debugging. (2) `SEND` needs to pass two offsets to the generator: one for
|
|
|
|
`RETURN` and one for `YIELD`. It uses the `oparg` for one, and the
|
|
|
|
`return_offset` for the other.
|