[3.13] gh-113993: Allow interned strings to be mortal, and fix related issues (GH-120520) (GH-120945)

* Add an InternalDocs file describing how interning should work and how to use it.

* Add internal functions to *explicitly* request what kind of interning is done:
  - `_PyUnicode_InternMortal`
  - `_PyUnicode_InternImmortal`
  - `_PyUnicode_InternStatic`

* Switch uses of `PyUnicode_InternInPlace` to those.

* Disallow using `_Py_SetImmortal` on strings directly.
  You should use `_PyUnicode_InternImmortal` instead:
  - Strings should be interned before immortalization, otherwise you're possibly
    interning a immortalizing copy.
  - `_Py_SetImmortal` doesn't handle the `SSTATE_INTERNED_MORTAL` to
    `SSTATE_INTERNED_IMMORTAL` update, and those flags can't be changed in
    backports, as they are now part of public API and version-specific ABI.

* Add private `_only_immortal` argument for `sys.getunicodeinternedsize`, used in refleak test machinery.

* Make sure the statically allocated string singletons are unique. This means these sets are now disjoint:
  - `_Py_ID`
  - `_Py_STR` (including the empty string)
  - one-character latin-1 singletons

  Now, when you intern a singleton, that exact singleton will be interned.

* Add a `_Py_LATIN1_CHR` macro, use it instead of `_Py_ID`/`_Py_STR` for one-character latin-1 singletons everywhere (including Clinic).

* Intern `_Py_STR` singletons at startup.

* For free-threaded builds, intern `_Py_LATIN1_CHR` singletons at startup.

* Beef up the tests. Cover internal details (marked with `@cpython_only`).

* Add lots of assertions

Co-authored-by: Eric Snow <ericsnowcurrently@gmail.com>
This commit is contained in:
Petr Viktorin 2024-06-24 20:24:19 +02:00 committed by GitHub
parent 447e07ab3d
commit 9769b7ae06
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
42 changed files with 2460 additions and 1140 deletions

View file

@ -137,6 +137,7 @@ static PyObject *intern_one_constant(PyObject *op);
static int
intern_strings(PyObject *tuple)
{
PyInterpreterState *interp = _PyInterpreterState_GET();
Py_ssize_t i;
for (i = PyTuple_GET_SIZE(tuple); --i >= 0; ) {
@ -146,7 +147,7 @@ intern_strings(PyObject *tuple)
"non-string found in code slot");
return -1;
}
PyUnicode_InternInPlace(&_PyTuple_ITEMS(tuple)[i]);
_PyUnicode_InternMortal(interp, &_PyTuple_ITEMS(tuple)[i]);
}
return 0;
}
@ -157,12 +158,13 @@ intern_strings(PyObject *tuple)
static int
intern_constants(PyObject *tuple, int *modified)
{
PyInterpreterState *interp = _PyInterpreterState_GET();
for (Py_ssize_t i = PyTuple_GET_SIZE(tuple); --i >= 0; ) {
PyObject *v = PyTuple_GET_ITEM(tuple, i);
if (PyUnicode_CheckExact(v)) {
if (should_intern_string(v)) {
PyObject *w = v;
PyUnicode_InternInPlace(&v);
_PyUnicode_InternMortal(interp, &v);
if (w != v) {
PyTuple_SET_ITEM(tuple, i, v);
if (modified) {
@ -458,12 +460,13 @@ init_code(PyCodeObject *co, struct _PyCodeConstructor *con)
con->stacksize = 1;
}
PyInterpreterState *interp = _PyInterpreterState_GET();
co->co_filename = Py_NewRef(con->filename);
co->co_name = Py_NewRef(con->name);
co->co_qualname = Py_NewRef(con->qualname);
PyUnicode_InternInPlace(&co->co_filename);
PyUnicode_InternInPlace(&co->co_name);
PyUnicode_InternInPlace(&co->co_qualname);
_PyUnicode_InternMortal(interp, &co->co_filename);
_PyUnicode_InternMortal(interp, &co->co_name);
_PyUnicode_InternMortal(interp, &co->co_qualname);
co->co_flags = con->flags;
co->co_firstlineno = con->firstlineno;
@ -489,7 +492,6 @@ init_code(PyCodeObject *co, struct _PyCodeConstructor *con)
co->co_framesize = nlocalsplus + con->stacksize + FRAME_SPECIALS_SIZE;
co->co_ncellvars = ncellvars;
co->co_nfreevars = nfreevars;
PyInterpreterState *interp = _PyInterpreterState_GET();
#ifdef Py_GIL_DISABLED
PyMutex_Lock(&interp->func_state.mutex);
#endif