cpython/Include/internal/pycore_optimizer.h
Guido van Rossum 3409bc29c9
gh-115859: Re-enable T2 optimizer pass by default (#116062)
This undoes the *temporary* default disabling of the T2 optimizer pass in gh-115860.

- Add a new test that reproduces Brandt's example from gh-115859; it indeed crashes before gh-116028 with PYTHONUOPSOPTIMIZE=1
- Re-enable the optimizer pass in T2, stop checking PYTHONUOPSOPTIMIZE
- Rename the env var to disable T2 entirely to PYTHON_UOPS_OPTIMIZE (must be explicitly set to 0 to disable)
- Fix skipIf conditions on tests in test_opt.py accordingly
- Export sym_is_bottom() (for debugging)
- Fix various things in the `_BINARY_OP_` specializations in the abstract interpreter:
  - DECREF(temp)
  - out-of-space check after sym_new_const()
  - add sym_matches_type() checks, so even if we somehow reach a binary op with symbolic constants of the wrong type on the stack we won't trigger the type assert
2024-02-28 22:38:01 +00:00

117 lines
3.6 KiB
C

#ifndef Py_INTERNAL_OPTIMIZER_H
#define Py_INTERNAL_OPTIMIZER_H
#ifdef __cplusplus
extern "C" {
#endif
#ifndef Py_BUILD_CORE
# error "this header requires Py_BUILD_CORE define"
#endif
#include "pycore_uop_ids.h"
#include <stdbool.h>
// This is the length of the trace we project initially.
#define UOP_MAX_TRACE_LENGTH 512
#define TRACE_STACK_SIZE 5
int _Py_uop_analyze_and_optimize(_PyInterpreterFrame *frame,
_PyUOpInstruction *trace, int trace_len, int curr_stackentries,
_PyBloomFilter *dependencies);
extern PyTypeObject _PyCounterExecutor_Type;
extern PyTypeObject _PyCounterOptimizer_Type;
extern PyTypeObject _PyDefaultOptimizer_Type;
extern PyTypeObject _PyUOpExecutor_Type;
extern PyTypeObject _PyUOpOptimizer_Type;
/* Symbols */
/* See explanation in optimizer_symbols.c */
struct _Py_UopsSymbol {
int flags; // 0 bits: Top; 2 or more bits: Bottom
PyTypeObject *typ; // Borrowed reference
PyObject *const_val; // Owned reference (!)
};
// Holds locals, stack, locals, stack ... co_consts (in that order)
#define MAX_ABSTRACT_INTERP_SIZE 4096
#define TY_ARENA_SIZE (UOP_MAX_TRACE_LENGTH * 5)
// Need extras for root frame and for overflow frame (see TRACE_STACK_PUSH())
#define MAX_ABSTRACT_FRAME_DEPTH (TRACE_STACK_SIZE + 2)
typedef struct _Py_UopsSymbol _Py_UopsSymbol;
struct _Py_UOpsAbstractFrame {
// Max stacklen
int stack_len;
int locals_len;
_Py_UopsSymbol **stack_pointer;
_Py_UopsSymbol **stack;
_Py_UopsSymbol **locals;
};
typedef struct _Py_UOpsAbstractFrame _Py_UOpsAbstractFrame;
typedef struct ty_arena {
int ty_curr_number;
int ty_max_number;
_Py_UopsSymbol arena[TY_ARENA_SIZE];
} ty_arena;
struct _Py_UOpsContext {
PyObject_HEAD
// The current "executing" frame.
_Py_UOpsAbstractFrame *frame;
_Py_UOpsAbstractFrame frames[MAX_ABSTRACT_FRAME_DEPTH];
int curr_frame_depth;
// Arena for the symbolic types.
ty_arena t_arena;
_Py_UopsSymbol **n_consumed;
_Py_UopsSymbol **limit;
_Py_UopsSymbol *locals_and_stack[MAX_ABSTRACT_INTERP_SIZE];
};
typedef struct _Py_UOpsContext _Py_UOpsContext;
extern bool _Py_uop_sym_is_null(_Py_UopsSymbol *sym);
extern bool _Py_uop_sym_is_not_null(_Py_UopsSymbol *sym);
extern bool _Py_uop_sym_is_const(_Py_UopsSymbol *sym);
extern PyObject *_Py_uop_sym_get_const(_Py_UopsSymbol *sym);
extern _Py_UopsSymbol *_Py_uop_sym_new_unknown(_Py_UOpsContext *ctx);
extern _Py_UopsSymbol *_Py_uop_sym_new_not_null(_Py_UOpsContext *ctx);
extern _Py_UopsSymbol *_Py_uop_sym_new_type(
_Py_UOpsContext *ctx, PyTypeObject *typ);
extern _Py_UopsSymbol *_Py_uop_sym_new_const(_Py_UOpsContext *ctx, PyObject *const_val);
extern _Py_UopsSymbol *_Py_uop_sym_new_null(_Py_UOpsContext *ctx);
extern bool _Py_uop_sym_matches_type(_Py_UopsSymbol *sym, PyTypeObject *typ);
extern void _Py_uop_sym_set_null(_Py_UopsSymbol *sym);
extern void _Py_uop_sym_set_non_null(_Py_UopsSymbol *sym);
extern void _Py_uop_sym_set_type(_Py_UopsSymbol *sym, PyTypeObject *typ);
extern void _Py_uop_sym_set_const(_Py_UopsSymbol *sym, PyObject *const_val);
extern bool _Py_uop_sym_is_bottom(_Py_UopsSymbol *sym);
extern int _Py_uop_abstractcontext_init(_Py_UOpsContext *ctx);
extern void _Py_uop_abstractcontext_fini(_Py_UOpsContext *ctx);
extern _Py_UOpsAbstractFrame *_Py_uop_frame_new(
_Py_UOpsContext *ctx,
PyCodeObject *co,
_Py_UopsSymbol **localsplus_start,
int n_locals_already_filled,
int curr_stackentries);
extern int _Py_uop_frame_pop(_Py_UOpsContext *ctx);
PyAPI_FUNC(PyObject *) _Py_uop_symbols_test(PyObject *self, PyObject *ignored);
#ifdef __cplusplus
}
#endif
#endif /* !Py_INTERNAL_OPTIMIZER_H */