We need an API for optimizers to be plugged in to CPython.
The proposed model is that of client server, where the VM is the client and the optimizer is the server.
The optimizer registers with the VM, then VM calls the optimizer when hotspots are detected.
The API:
type struct {
OBJECT_HEADER;
_PyInterpreterFrame *(*execute)(PyExecutorObject *self, _PyInterpreterFrame *frame, PyObject **stack_pointer);
/* Data needed by the executor goes here, but is opaque to the VM */
} PyExecutorObject;
/* This would be nicer as an enum, but C doesn't define the size of enums */
#definePY_OPTIMIZE_FUNCTION_ENTRY1
#definePY_OPTIMIZE_RESUME_AFTER_YIELD2
#definePY_OPTIMIZE_BACK_EDGE4typedefuint32_t PyOptimizerCapabilities;
type struct {
OBJECT_HEADER;
PyExecutorObject *(*compile)(PyOptimizerObject* self, PyCodeObject *code, int offset);
PyOptimizerCapabilities capabilities;
float optimization_cost;
float run_cost;
/* Data needed by the compiler goes here, but is opaque to the VM */
} PyOptimizerObject;
void_Py_Executor_Replace(PyCodeObject *code, int offset, PyExecutorObject *executor);
int_Py_Optimizer_Register(PyOptimizerObject* optimizer);
The semantics of a PyExecutorObject is that upon return from its execute function, the VM state will have advanced N instructions. Where N is a non-negative integer.
Note that the above API is just the initial version to support our work on speeding up Python 3.13.
It will probably need to extended to support PyTorch Dynamo and other users of PEP 523 that cannot use PEP 669, but that is for another issue.
We need an API for optimizers to be plugged in to CPython.
The proposed model is that of client server, where the VM is the client and the optimizer is the server.
The optimizer registers with the VM, then VM calls the optimizer when hotspots are detected.
The API:
The semantics of a
PyExecutorObject
is that upon return from itsexecute
function, the VM state will have advancedN
instructions. WhereN
is a non-negative integer.Full discussion here: faster-cpython/ideas#380
This is not a replacement for PEP 523. That will need a PEP. We should get this working first, before we consider replacing PEP 523.
Linked PRs
The text was updated successfully, but these errors were encountered: