gh-106931: Intern Statically Allocated Strings Globally (gh-107272)

We tried this before with a dict and for all interned strings.  That ran into problems due to interpreter isolation.  However, exclusively using a per-interpreter cache caused some inconsistency that can eliminate the benefit of interning.  Here we circle back to using a global cache, but only for statically allocated strings.  We also use a more-basic _Py_hashtable_t for that global cache instead of a dict.

Ideally we would only have the global cache, but the optional isolation of each interpreter's allocator means that a non-static string object must not outlive its interpreter.  Thus we would have to store a copy of each such interned string in the global cache, tied to the main interpreter.

4 files changed, 9 insertions, 0 deletions

diff --git a/Include/internal/pycore_global_objects.h b/Include/internal/pycore_global_objects.h
index 5a3fb13..442f851 100644
--- a/Include/internal/pycore_global_objects.h
+++ b/Include/internal/pycore_global_objects.h
@@ -8,6 +8,7 @@ extern "C" {
 #  error "this header requires Py_BUILD_CORE define"
 #endif
 
+#include "pycore_hashtable.h"       // _Py_hashtable_t
 #include "pycore_gc.h"              // PyGC_Head
 #include "pycore_global_strings.h"  // struct _Py_global_strings
 #include "pycore_hamt.h"            // PyHamtNode_Bitmap
@@ -28,6 +29,11 @@ extern "C" {
 #define _Py_SINGLETON(NAME) \
     _Py_GLOBAL_OBJECT(singletons.NAME)
 
+struct _Py_cached_objects {
+    // XXX We could statically allocate the hashtable.
+    _Py_hashtable_t *interned_strings;
+};
+
 struct _Py_static_objects {
     struct {
         /* Small integers are preallocated in this array so that they
diff --git a/Include/internal/pycore_hashtable.h b/Include/internal/pycore_hashtable.h
index 6501ab1..f57978a 100644
--- a/Include/internal/pycore_hashtable.h
+++ b/Include/internal/pycore_hashtable.h
@@ -106,6 +106,7 @@ PyAPI_FUNC(int) _Py_hashtable_foreach(
     void *user_data);
 
 PyAPI_FUNC(size_t) _Py_hashtable_size(const _Py_hashtable_t *ht);
+PyAPI_FUNC(size_t) _Py_hashtable_len(const _Py_hashtable_t *ht);
 
 /* Add a new entry to the hash. The key must not be present in the hash table.
    Return 0 on success, -1 on memory error. */
diff --git a/Include/internal/pycore_runtime.h b/Include/internal/pycore_runtime.h
index eb6b66b..0ec86ee 100644
--- a/Include/internal/pycore_runtime.h
+++ b/Include/internal/pycore_runtime.h
@@ -249,6 +249,7 @@ typedef struct pyruntimestate {
     struct _types_runtime_state types;
 
     /* All the objects that are shared by the runtime's interpreters. */
+    struct _Py_cached_objects cached_objects;
     struct _Py_static_objects static_objects;
 
     /* The following fields are here to avoid allocation during init.
diff --git a/Include/internal/pycore_runtime_init.h b/Include/internal/pycore_runtime_init.h
index e72e742..840e41c 100644
--- a/Include/internal/pycore_runtime_init.h
+++ b/Include/internal/pycore_runtime_init.h
@@ -214,6 +214,7 @@ extern PyTypeObject _PyExc_MemoryError;
             .kind = 1, \
             .compact = 1, \
             .ascii = (ASCII), \
+            .statically_allocated = 1, \
         }, \
     }
 #define _PyASCIIObject_INIT(LITERAL) \