gh-114940: Add _Py_FOR_EACH_TSTATE_UNLOCKED(), and Friends (gh-127077)

This is a precursor to the actual fix for gh-114940, where we will change these macros to use the new lock.  This change is almost entirely mechanical; the exceptions are the loops in codeobject.c and ceval.c, which now hold the "head" lock.  Note that almost all of the uses of _Py_FOR_EACH_TSTATE_UNLOCKED() here will change to _Py_FOR_EACH_TSTATE_BEGIN() once we add the new per-interpreter lock.

3 files changed, 6 insertions, 4 deletions

diff --git a/Objects/codeobject.c b/Objects/codeobject.c
index ec5d8a5..148350c 100644
--- a/Objects/codeobject.c
+++ b/Objects/codeobject.c
@@ -2895,20 +2895,22 @@ get_indices_in_use(PyInterpreterState *interp, struct flag_set *in_use)
     assert(interp->stoptheworld.world_stopped);
     assert(in_use->flags == NULL);
     int32_t max_index = 0;
-    for (PyThreadState *p = interp->threads.head; p != NULL; p = p->next) {
+    _Py_FOR_EACH_TSTATE_BEGIN(interp, p) {
         int32_t idx = ((_PyThreadStateImpl *) p)->tlbc_index;
         if (idx > max_index) {
             max_index = idx;
         }
     }
+    _Py_FOR_EACH_TSTATE_END(interp);
     in_use->size = (size_t) max_index + 1;
     in_use->flags = PyMem_Calloc(in_use->size, sizeof(*in_use->flags));
     if (in_use->flags == NULL) {
         return -1;
     }
-    for (PyThreadState *p = interp->threads.head; p != NULL; p = p->next) {
+    _Py_FOR_EACH_TSTATE_BEGIN(interp, p) {
         in_use->flags[((_PyThreadStateImpl *) p)->tlbc_index] = 1;
     }
+    _Py_FOR_EACH_TSTATE_END(interp);
     return 0;
 }
 
diff --git a/Objects/object.c b/Objects/object.c
index b115bc7..8868fa2 100644
--- a/Objects/object.c
+++ b/Objects/object.c
@@ -119,7 +119,7 @@ get_reftotal(PyInterpreterState *interp)
        since we can't determine which interpreter updated it. */
     Py_ssize_t total = REFTOTAL(interp);
 #ifdef Py_GIL_DISABLED
-    for (PyThreadState *p = interp->threads.head; p != NULL; p = p->next) {
+    _Py_FOR_EACH_TSTATE_UNLOCKED(interp, p) {
         /* This may race with other threads modifications to their reftotal */
         _PyThreadStateImpl *tstate_impl = (_PyThreadStateImpl *)p;
         total += _Py_atomic_load_ssize_relaxed(&tstate_impl->reftotal);
diff --git a/Objects/obmalloc.c b/Objects/obmalloc.c
index 3d6b1ab..2cc0377 100644
--- a/Objects/obmalloc.c
+++ b/Objects/obmalloc.c
@@ -1439,7 +1439,7 @@ get_mimalloc_allocated_blocks(PyInterpreterState *interp)
 {
     size_t allocated_blocks = 0;
 #ifdef Py_GIL_DISABLED
-    for (PyThreadState *t = interp->threads.head; t != NULL; t = t->next) {
+    _Py_FOR_EACH_TSTATE_UNLOCKED(interp, t) {
         _PyThreadStateImpl *tstate = (_PyThreadStateImpl *)t;
         for (int i = 0; i < _Py_MIMALLOC_HEAP_COUNT; i++) {
             mi_heap_t *heap = &tstate->mimalloc.heaps[i];