bpo-45953: Statically allocate and initialize global bytes objects. (gh-30096)

The empty bytes object (b'') and the 256 one-character bytes objects were allocated at runtime init.  Now we statically allocate and initialize them.

https://bugs.python.org/issue45953

5 files changed, 294 insertions, 96 deletions

diff --git a/Include/internal/pycore_bytesobject.h b/Include/internal/pycore_bytesobject.h
index b00ed97..18d9530 100644
--- a/Include/internal/pycore_bytesobject.h
+++ b/Include/internal/pycore_bytesobject.h
@@ -11,17 +11,7 @@ extern "C" {
 
 /* runtime lifecycle */
 
-extern PyStatus _PyBytes_InitGlobalObjects(PyInterpreterState *);
 extern PyStatus _PyBytes_InitTypes(PyInterpreterState *);
-extern void _PyBytes_Fini(PyInterpreterState *);
-
-
-/* other API */
-
-struct _Py_bytes_state {
-    PyObject *empty_string;
-    PyBytesObject *characters[256];
-};
 
 
 #ifdef __cplusplus
diff --git a/Include/internal/pycore_global_objects.h b/Include/internal/pycore_global_objects.h
index 6cae3bc..d2dc907 100644
--- a/Include/internal/pycore_global_objects.h
+++ b/Include/internal/pycore_global_objects.h
@@ -34,6 +34,20 @@ extern "C" {
     }
 
 
+/* bytes objects */
+
+#define _PyBytes_SIMPLE_INIT(CH, LEN) \
+    { \
+        _PyVarObject_IMMORTAL_INIT(&PyBytes_Type, LEN), \
+        .ob_shash = -1, \
+        .ob_sval = { CH }, \
+    }
+#define _PyBytes_CHAR_INIT(CH) \
+    { \
+        _PyBytes_SIMPLE_INIT(CH, 1) \
+    }
+
+
 /**********************
  * the global objects *
  **********************/
@@ -54,6 +68,12 @@ struct _Py_global_objects {
          * -_PY_NSMALLNEGINTS (inclusive) to _PY_NSMALLPOSINTS (exclusive).
          */
         PyLongObject small_ints[_PY_NSMALLNEGINTS + _PY_NSMALLPOSINTS];
+
+        PyBytesObject bytes_empty;
+        struct {
+            PyBytesObject ob;
+            char eos;
+        } bytes_characters[256];
     } singletons;
 };
 
@@ -323,6 +343,266 @@ struct _Py_global_objects {
             _PyLong_DIGIT_INIT(255), \
             _PyLong_DIGIT_INIT(256), \
         }, \
+        \
+        .bytes_empty = _PyBytes_SIMPLE_INIT(0, 0), \
+        .bytes_characters = { \
+            _PyBytes_CHAR_INIT(0), \
+            _PyBytes_CHAR_INIT(1), \
+            _PyBytes_CHAR_INIT(2), \
+            _PyBytes_CHAR_INIT(3), \
+            _PyBytes_CHAR_INIT(4), \
+            _PyBytes_CHAR_INIT(5), \
+            _PyBytes_CHAR_INIT(6), \
+            _PyBytes_CHAR_INIT(7), \
+            _PyBytes_CHAR_INIT(8), \
+            _PyBytes_CHAR_INIT(9), \
+            _PyBytes_CHAR_INIT(10), \
+            _PyBytes_CHAR_INIT(11), \
+            _PyBytes_CHAR_INIT(12), \
+            _PyBytes_CHAR_INIT(13), \
+            _PyBytes_CHAR_INIT(14), \
+            _PyBytes_CHAR_INIT(15), \
+            _PyBytes_CHAR_INIT(16), \
+            _PyBytes_CHAR_INIT(17), \
+            _PyBytes_CHAR_INIT(18), \
+            _PyBytes_CHAR_INIT(19), \
+            _PyBytes_CHAR_INIT(20), \
+            _PyBytes_CHAR_INIT(21), \
+            _PyBytes_CHAR_INIT(22), \
+            _PyBytes_CHAR_INIT(23), \
+            _PyBytes_CHAR_INIT(24), \
+            _PyBytes_CHAR_INIT(25), \
+            _PyBytes_CHAR_INIT(26), \
+            _PyBytes_CHAR_INIT(27), \
+            _PyBytes_CHAR_INIT(28), \
+            _PyBytes_CHAR_INIT(29), \
+            _PyBytes_CHAR_INIT(30), \
+            _PyBytes_CHAR_INIT(31), \
+            _PyBytes_CHAR_INIT(32), \
+            _PyBytes_CHAR_INIT(33), \
+            _PyBytes_CHAR_INIT(34), \
+            _PyBytes_CHAR_INIT(35), \
+            _PyBytes_CHAR_INIT(36), \
+            _PyBytes_CHAR_INIT(37), \
+            _PyBytes_CHAR_INIT(38), \
+            _PyBytes_CHAR_INIT(39), \
+            _PyBytes_CHAR_INIT(40), \
+            _PyBytes_CHAR_INIT(41), \
+            _PyBytes_CHAR_INIT(42), \
+            _PyBytes_CHAR_INIT(43), \
+            _PyBytes_CHAR_INIT(44), \
+            _PyBytes_CHAR_INIT(45), \
+            _PyBytes_CHAR_INIT(46), \
+            _PyBytes_CHAR_INIT(47), \
+            _PyBytes_CHAR_INIT(48), \
+            _PyBytes_CHAR_INIT(49), \
+            _PyBytes_CHAR_INIT(50), \
+            _PyBytes_CHAR_INIT(51), \
+            _PyBytes_CHAR_INIT(52), \
+            _PyBytes_CHAR_INIT(53), \
+            _PyBytes_CHAR_INIT(54), \
+            _PyBytes_CHAR_INIT(55), \
+            _PyBytes_CHAR_INIT(56), \
+            _PyBytes_CHAR_INIT(57), \
+            _PyBytes_CHAR_INIT(58), \
+            _PyBytes_CHAR_INIT(59), \
+            _PyBytes_CHAR_INIT(60), \
+            _PyBytes_CHAR_INIT(61), \
+            _PyBytes_CHAR_INIT(62), \
+            _PyBytes_CHAR_INIT(63), \
+            _PyBytes_CHAR_INIT(64), \
+            _PyBytes_CHAR_INIT(65), \
+            _PyBytes_CHAR_INIT(66), \
+            _PyBytes_CHAR_INIT(67), \
+            _PyBytes_CHAR_INIT(68), \
+            _PyBytes_CHAR_INIT(69), \
+            _PyBytes_CHAR_INIT(70), \
+            _PyBytes_CHAR_INIT(71), \
+            _PyBytes_CHAR_INIT(72), \
+            _PyBytes_CHAR_INIT(73), \
+            _PyBytes_CHAR_INIT(74), \
+            _PyBytes_CHAR_INIT(75), \
+            _PyBytes_CHAR_INIT(76), \
+            _PyBytes_CHAR_INIT(77), \
+            _PyBytes_CHAR_INIT(78), \
+            _PyBytes_CHAR_INIT(79), \
+            _PyBytes_CHAR_INIT(80), \
+            _PyBytes_CHAR_INIT(81), \
+            _PyBytes_CHAR_INIT(82), \
+            _PyBytes_CHAR_INIT(83), \
+            _PyBytes_CHAR_INIT(84), \
+            _PyBytes_CHAR_INIT(85), \
+            _PyBytes_CHAR_INIT(86), \
+            _PyBytes_CHAR_INIT(87), \
+            _PyBytes_CHAR_INIT(88), \
+            _PyBytes_CHAR_INIT(89), \
+            _PyBytes_CHAR_INIT(90), \
+            _PyBytes_CHAR_INIT(91), \
+            _PyBytes_CHAR_INIT(92), \
+            _PyBytes_CHAR_INIT(93), \
+            _PyBytes_CHAR_INIT(94), \
+            _PyBytes_CHAR_INIT(95), \
+            _PyBytes_CHAR_INIT(96), \
+            _PyBytes_CHAR_INIT(97), \
+            _PyBytes_CHAR_INIT(98), \
+            _PyBytes_CHAR_INIT(99), \
+            _PyBytes_CHAR_INIT(100), \
+            _PyBytes_CHAR_INIT(101), \
+            _PyBytes_CHAR_INIT(102), \
+            _PyBytes_CHAR_INIT(103), \
+            _PyBytes_CHAR_INIT(104), \
+            _PyBytes_CHAR_INIT(105), \
+            _PyBytes_CHAR_INIT(106), \
+            _PyBytes_CHAR_INIT(107), \
+            _PyBytes_CHAR_INIT(108), \
+            _PyBytes_CHAR_INIT(109), \
+            _PyBytes_CHAR_INIT(110), \
+            _PyBytes_CHAR_INIT(111), \
+            _PyBytes_CHAR_INIT(112), \
+            _PyBytes_CHAR_INIT(113), \
+            _PyBytes_CHAR_INIT(114), \
+            _PyBytes_CHAR_INIT(115), \
+            _PyBytes_CHAR_INIT(116), \
+            _PyBytes_CHAR_INIT(117), \
+            _PyBytes_CHAR_INIT(118), \
+            _PyBytes_CHAR_INIT(119), \
+            _PyBytes_CHAR_INIT(120), \
+            _PyBytes_CHAR_INIT(121), \
+            _PyBytes_CHAR_INIT(122), \
+            _PyBytes_CHAR_INIT(123), \
+            _PyBytes_CHAR_INIT(124), \
+            _PyBytes_CHAR_INIT(125), \
+            _PyBytes_CHAR_INIT(126), \
+            _PyBytes_CHAR_INIT(127), \
+            _PyBytes_CHAR_INIT(128), \
+            _PyBytes_CHAR_INIT(129), \
+            _PyBytes_CHAR_INIT(130), \
+            _PyBytes_CHAR_INIT(131), \
+            _PyBytes_CHAR_INIT(132), \
+            _PyBytes_CHAR_INIT(133), \
+            _PyBytes_CHAR_INIT(134), \
+            _PyBytes_CHAR_INIT(135), \
+            _PyBytes_CHAR_INIT(136), \
+            _PyBytes_CHAR_INIT(137), \
+            _PyBytes_CHAR_INIT(138), \
+            _PyBytes_CHAR_INIT(139), \
+            _PyBytes_CHAR_INIT(140), \
+            _PyBytes_CHAR_INIT(141), \
+            _PyBytes_CHAR_INIT(142), \
+            _PyBytes_CHAR_INIT(143), \
+            _PyBytes_CHAR_INIT(144), \
+            _PyBytes_CHAR_INIT(145), \
+            _PyBytes_CHAR_INIT(146), \
+            _PyBytes_CHAR_INIT(147), \
+            _PyBytes_CHAR_INIT(148), \
+            _PyBytes_CHAR_INIT(149), \
+            _PyBytes_CHAR_INIT(150), \
+            _PyBytes_CHAR_INIT(151), \
+            _PyBytes_CHAR_INIT(152), \
+            _PyBytes_CHAR_INIT(153), \
+            _PyBytes_CHAR_INIT(154), \
+            _PyBytes_CHAR_INIT(155), \
+            _PyBytes_CHAR_INIT(156), \
+            _PyBytes_CHAR_INIT(157), \
+            _PyBytes_CHAR_INIT(158), \
+            _PyBytes_CHAR_INIT(159), \
+            _PyBytes_CHAR_INIT(160), \
+            _PyBytes_CHAR_INIT(161), \
+            _PyBytes_CHAR_INIT(162), \
+            _PyBytes_CHAR_INIT(163), \
+            _PyBytes_CHAR_INIT(164), \
+            _PyBytes_CHAR_INIT(165), \
+            _PyBytes_CHAR_INIT(166), \
+            _PyBytes_CHAR_INIT(167), \
+            _PyBytes_CHAR_INIT(168), \
+            _PyBytes_CHAR_INIT(169), \
+            _PyBytes_CHAR_INIT(170), \
+            _PyBytes_CHAR_INIT(171), \
+            _PyBytes_CHAR_INIT(172), \
+            _PyBytes_CHAR_INIT(173), \
+            _PyBytes_CHAR_INIT(174), \
+            _PyBytes_CHAR_INIT(175), \
+            _PyBytes_CHAR_INIT(176), \
+            _PyBytes_CHAR_INIT(177), \
+            _PyBytes_CHAR_INIT(178), \
+            _PyBytes_CHAR_INIT(179), \
+            _PyBytes_CHAR_INIT(180), \
+            _PyBytes_CHAR_INIT(181), \
+            _PyBytes_CHAR_INIT(182), \
+            _PyBytes_CHAR_INIT(183), \
+            _PyBytes_CHAR_INIT(184), \
+            _PyBytes_CHAR_INIT(185), \
+            _PyBytes_CHAR_INIT(186), \
+            _PyBytes_CHAR_INIT(187), \
+            _PyBytes_CHAR_INIT(188), \
+            _PyBytes_CHAR_INIT(189), \
+            _PyBytes_CHAR_INIT(190), \
+            _PyBytes_CHAR_INIT(191), \
+            _PyBytes_CHAR_INIT(192), \
+            _PyBytes_CHAR_INIT(193), \
+            _PyBytes_CHAR_INIT(194), \
+            _PyBytes_CHAR_INIT(195), \
+            _PyBytes_CHAR_INIT(196), \
+            _PyBytes_CHAR_INIT(197), \
+            _PyBytes_CHAR_INIT(198), \
+            _PyBytes_CHAR_INIT(199), \
+            _PyBytes_CHAR_INIT(200), \
+            _PyBytes_CHAR_INIT(201), \
+            _PyBytes_CHAR_INIT(202), \
+            _PyBytes_CHAR_INIT(203), \
+            _PyBytes_CHAR_INIT(204), \
+            _PyBytes_CHAR_INIT(205), \
+            _PyBytes_CHAR_INIT(206), \
+            _PyBytes_CHAR_INIT(207), \
+            _PyBytes_CHAR_INIT(208), \
+            _PyBytes_CHAR_INIT(209), \
+            _PyBytes_CHAR_INIT(210), \
+            _PyBytes_CHAR_INIT(211), \
+            _PyBytes_CHAR_INIT(212), \
+            _PyBytes_CHAR_INIT(213), \
+            _PyBytes_CHAR_INIT(214), \
+            _PyBytes_CHAR_INIT(215), \
+            _PyBytes_CHAR_INIT(216), \
+            _PyBytes_CHAR_INIT(217), \
+            _PyBytes_CHAR_INIT(218), \
+            _PyBytes_CHAR_INIT(219), \
+            _PyBytes_CHAR_INIT(220), \
+            _PyBytes_CHAR_INIT(221), \
+            _PyBytes_CHAR_INIT(222), \
+            _PyBytes_CHAR_INIT(223), \
+            _PyBytes_CHAR_INIT(224), \
+            _PyBytes_CHAR_INIT(225), \
+            _PyBytes_CHAR_INIT(226), \
+            _PyBytes_CHAR_INIT(227), \
+            _PyBytes_CHAR_INIT(228), \
+            _PyBytes_CHAR_INIT(229), \
+            _PyBytes_CHAR_INIT(230), \
+            _PyBytes_CHAR_INIT(231), \
+            _PyBytes_CHAR_INIT(232), \
+            _PyBytes_CHAR_INIT(233), \
+            _PyBytes_CHAR_INIT(234), \
+            _PyBytes_CHAR_INIT(235), \
+            _PyBytes_CHAR_INIT(236), \
+            _PyBytes_CHAR_INIT(237), \
+            _PyBytes_CHAR_INIT(238), \
+            _PyBytes_CHAR_INIT(239), \
+            _PyBytes_CHAR_INIT(240), \
+            _PyBytes_CHAR_INIT(241), \
+            _PyBytes_CHAR_INIT(242), \
+            _PyBytes_CHAR_INIT(243), \
+            _PyBytes_CHAR_INIT(244), \
+            _PyBytes_CHAR_INIT(245), \
+            _PyBytes_CHAR_INIT(246), \
+            _PyBytes_CHAR_INIT(247), \
+            _PyBytes_CHAR_INIT(248), \
+            _PyBytes_CHAR_INIT(249), \
+            _PyBytes_CHAR_INIT(250), \
+            _PyBytes_CHAR_INIT(251), \
+            _PyBytes_CHAR_INIT(252), \
+            _PyBytes_CHAR_INIT(253), \
+            _PyBytes_CHAR_INIT(254), \
+            _PyBytes_CHAR_INIT(255), \
+        }, \
     }, \
 }
 
diff --git a/Include/internal/pycore_interp.h b/Include/internal/pycore_interp.h
index e4d7b1b..d48ea87 100644
--- a/Include/internal/pycore_interp.h
+++ b/Include/internal/pycore_interp.h
@@ -10,7 +10,6 @@ extern "C" {
 
 #include "pycore_atomic.h"        // _Py_atomic_address
 #include "pycore_ast_state.h"     // struct ast_state
-#include "pycore_bytesobject.h"   // struct _Py_bytes_state
 #include "pycore_context.h"       // struct _Py_context_state
 #include "pycore_dict.h"          // struct _Py_dict_state
 #include "pycore_exceptions.h"    // struct _Py_exc_state
@@ -152,7 +151,6 @@ struct _is {
 
     PyObject *audit_hooks;
 
-    struct _Py_bytes_state bytes;
     struct _Py_unicode_state unicode;
     struct _Py_float_state float_state;
     /* Using a cache is very effective since typically only a single slice is
diff --git a/Objects/bytesobject.c b/Objects/bytesobject.c
index 2f7e0a6..85d6912 100644
--- a/Objects/bytesobject.c
+++ b/Objects/bytesobject.c
@@ -5,9 +5,9 @@
 #include "Python.h"
 #include "pycore_abstract.h"      // _PyIndex_Check()
 #include "pycore_bytes_methods.h" // _Py_bytes_startswith()
-#include "pycore_bytesobject.h"   // struct _Py_bytes_state
 #include "pycore_call.h"          // _PyObject_CallNoArgs()
 #include "pycore_format.h"        // F_LJUST
+#include "pycore_global_objects.h"  // _Py_GET_GLOBAL_OBJECT()
 #include "pycore_initconfig.h"    // _PyStatus_OK()
 #include "pycore_long.h"          // _PyLong_DigitValue
 #include "pycore_object.h"        // _PyObject_GC_TRACK
@@ -38,49 +38,24 @@ Py_LOCAL_INLINE(Py_ssize_t) _PyBytesWriter_GetSize(_PyBytesWriter *writer,
                                                    char *str);
 
 
-static struct _Py_bytes_state*
-get_bytes_state(void)
-{
-    PyInterpreterState *interp = _PyInterpreterState_GET();
-    return &interp->bytes;
-}
+#define CHARACTERS _Py_SINGLETON(bytes_characters)
+#define CHARACTER(ch) \
+     ((PyBytesObject *)&(CHARACTERS[ch]));
+#define EMPTY (&_Py_SINGLETON(bytes_empty))
 
 
 // Return a borrowed reference to the empty bytes string singleton.
 static inline PyObject* bytes_get_empty(void)
 {
-    struct _Py_bytes_state *state = get_bytes_state();
-    // bytes_get_empty() must not be called before _PyBytes_Init()
-    // or after _PyBytes_Fini()
-    assert(state->empty_string != NULL);
-    return state->empty_string;
+    return &EMPTY->ob_base.ob_base;
 }
 
 
 // Return a strong reference to the empty bytes string singleton.
 static inline PyObject* bytes_new_empty(void)
 {
-    PyObject *empty = bytes_get_empty();
-    Py_INCREF(empty);
-    return (PyObject *)empty;
-}
-
-
-static int
-bytes_create_empty_string_singleton(struct _Py_bytes_state *state)
-{
-    // Create the empty bytes string singleton
-    PyBytesObject *op = (PyBytesObject *)PyObject_Malloc(PyBytesObject_SIZE);
-    if (op == NULL) {
-        return -1;
-    }
-    _PyObject_InitVar((PyVarObject*)op, &PyBytes_Type, 0);
-    op->ob_shash = -1;
-    op->ob_sval[0] = '\0';
-
-    assert(state->empty_string == NULL);
-    state->empty_string = (PyObject *)op;
-    return 0;
+    Py_INCREF(EMPTY);
+    return (PyObject *)EMPTY;
 }
 
 
@@ -148,12 +123,9 @@ PyBytes_FromStringAndSize(const char *str, Py_ssize_t size)
         return NULL;
     }
     if (size == 1 && str != NULL) {
-        struct _Py_bytes_state *state = get_bytes_state();
-        op = state->characters[*str & UCHAR_MAX];
-        if (op != NULL) {
-            Py_INCREF(op);
-            return (PyObject *)op;
-        }
+        op = CHARACTER(*str & 255);
+        Py_INCREF(op);
+        return (PyObject *)op;
     }
     if (size == 0) {
         return bytes_new_empty();
@@ -166,12 +138,6 @@ PyBytes_FromStringAndSize(const char *str, Py_ssize_t size)
         return (PyObject *) op;
 
     memcpy(op->ob_sval, str, size);
-    /* share short strings */
-    if (size == 1) {
-        struct _Py_bytes_state *state = get_bytes_state();
-        Py_INCREF(op);
-        state->characters[*str & UCHAR_MAX] = op;
-    }
     return (PyObject *) op;
 }
 
@@ -189,16 +155,13 @@ PyBytes_FromString(const char *str)
         return NULL;
     }
 
-    struct _Py_bytes_state *state = get_bytes_state();
     if (size == 0) {
         return bytes_new_empty();
     }
     else if (size == 1) {
-        op = state->characters[*str & UCHAR_MAX];
-        if (op != NULL) {
-            Py_INCREF(op);
-            return (PyObject *)op;
-        }
+        op = CHARACTER(*str & 255);
+        Py_INCREF(op);
+        return (PyObject *)op;
     }
 
     /* Inline PyObject_NewVar */
@@ -209,12 +172,6 @@ PyBytes_FromString(const char *str)
     _PyObject_InitVar((PyVarObject*)op, &PyBytes_Type, size);
     op->ob_shash = -1;
     memcpy(op->ob_sval, str, size+1);
-    /* share short strings */
-    if (size == 1) {
-        assert(state->characters[*str & UCHAR_MAX] == NULL);
-        Py_INCREF(op);
-        state->characters[*str & UCHAR_MAX] = op;
-    }
     return (PyObject *) op;
 }
 
@@ -3087,17 +3044,6 @@ error:
 
 
 PyStatus
-_PyBytes_InitGlobalObjects(PyInterpreterState *interp)
-{
-    struct _Py_bytes_state *state = &interp->bytes;
-    if (bytes_create_empty_string_singleton(state) < 0) {
-        return _PyStatus_NO_MEMORY();
-    }
-    return _PyStatus_OK();
-}
-
-
-PyStatus
 _PyBytes_InitTypes(PyInterpreterState *interp)
 {
     if (!_Py_IsMainInterpreter(interp)) {
@@ -3116,16 +3062,6 @@ _PyBytes_InitTypes(PyInterpreterState *interp)
 }
 
 
-void
-_PyBytes_Fini(PyInterpreterState *interp)
-{
-    struct _Py_bytes_state* state = &interp->bytes;
-    for (int i = 0; i < UCHAR_MAX + 1; i++) {
-        Py_CLEAR(state->characters[i]);
-    }
-    Py_CLEAR(state->empty_string);
-}
-
 /*********************** Bytes Iterator ****************************/
 
 typedef struct {
diff --git a/Python/pylifecycle.c b/Python/pylifecycle.c
index b2f58f4..284cfac 100644
--- a/Python/pylifecycle.c
+++ b/Python/pylifecycle.c
@@ -678,11 +678,6 @@ pycore_init_global_objects(PyInterpreterState *interp)
 
     _PyFloat_InitState(interp);
 
-    status = _PyBytes_InitGlobalObjects(interp);
-    if (_PyStatus_EXCEPTION(status)) {
-        return status;
-    }
-
     status = _PyUnicode_InitGlobalObjects(interp);
     if (_PyStatus_EXCEPTION(status)) {
         return status;
@@ -1685,7 +1680,6 @@ finalize_interp_types(PyInterpreterState *interp)
 
     _PySlice_Fini(interp);
 
-    _PyBytes_Fini(interp);
     _PyUnicode_Fini(interp);
     _PyFloat_Fini(interp);
 }