[3.10] bpo-45056: Remove trailing unused constants from co_consts (GH-28109) (GH-28125)

(cherry picked from commit 55c4a92)

Co-authored-by: Inada Naoki <songofacandy@gmail.com>

Author: ambv

Lib/test/test_compile.py

@@ -648,6 +648,17 @@ def test_merge_code_attrs(self):
         self.assertIs(f1.__code__.co_linetable, f2.__code__.co_linetable)
         self.assertIs(f1.__code__.co_code, f2.__code__.co_code)
 
+    # Stripping unused constants is not a strict requirement for the
+    # Python semantics, it's a more an implementation detail.
+    @support.cpython_only
+    def test_strip_unused_consts(self):
+        # Python 3.10rc1 appended None to co_consts when None is not used
+        # at all. See bpo-45056.
+        def f1():
+            "docstring"
+            return 42
+        self.assertEqual(f1.__code__.co_consts, ("docstring", 42))
+
     # This is a regression test for a CPython specific peephole optimizer
     # implementation bug present in a few releases.  It's assertion verifies
     # that peephole optimization was actually done though that isn't an

Lib/test/test_dis.py

@@ -689,10 +689,7 @@ def get_disassembly(self, func, lasti=-1, wrapper=True, **kwargs):
 if sys.flags.optimize:
     code_info_consts = "0: None"
 else:
-    code_info_consts = (
-    """0: 'Formatted details of methods, functions, or code.'
-   1: None"""
-)
+    code_info_consts = "0: 'Formatted details of methods, functions, or code.'"
 
 code_info_code_info = f"""\
 Name:              code_info
@@ -816,7 +813,6 @@ def f(c=c):
 Constants:
    0: 0
    1: 1
-   2: None
 Names:
    0: x"""
 

Misc/NEWS.d/next/Core and Builtins/2021-09-01-16-55-43.bpo-45056.7AK2d9.rst

@@ -0,0 +1 @@
+Compiler now removes trailing unused constants from co_consts.

Python/compile.c

@@ -6986,6 +6986,9 @@ normalize_basic_block(basicblock *bb);
 static int
 optimize_cfg(struct compiler *c, struct assembler *a, PyObject *consts);
 
+static int
+trim_unused_consts(struct compiler *c, struct assembler *a, PyObject *consts);
+
 /* Duplicates exit BBs, so that line numbers can be propagated to them */
 static int
 duplicate_exits_without_lineno(struct compiler *c);
@@ -7127,6 +7130,9 @@ assemble(struct compiler *c, int addNone)
     if (duplicate_exits_without_lineno(c)) {
         return NULL;
     }
+    if (trim_unused_consts(c, &a, consts)) {
+        goto error;
+    }
     propagate_line_numbers(&a);
     guarantee_lineno_for_exits(&a, c->u->u_firstlineno);
     /* Can't modify the bytecode after computing jump offsets. */
@@ -7809,6 +7815,33 @@ optimize_cfg(struct compiler *c, struct assembler *a, PyObject *consts)
     return 0;
 }
 
+// Remove trailing unused constants.
+static int
+trim_unused_consts(struct compiler *c, struct assembler *a, PyObject *consts)
+{
+    assert(PyList_CheckExact(consts));
+
+    // The first constant may be docstring; keep it always.
+    int max_const_index = 0;
+    for (basicblock *b = a->a_entry; b != NULL; b = b->b_next) {
+        for (int i = 0; i < b->b_iused; i++) {
+            if (b->b_instr[i].i_opcode == LOAD_CONST &&
+                    b->b_instr[i].i_oparg > max_const_index) {
+                max_const_index = b->b_instr[i].i_oparg;
+            }
+        }
+    }
+    if (max_const_index+1 < PyList_GET_SIZE(consts)) {
+        //fprintf(stderr, "removing trailing consts: max=%d, size=%d\n",
+        //        max_const_index, (int)PyList_GET_SIZE(consts));
+        if (PyList_SetSlice(consts, max_const_index+1,
+                            PyList_GET_SIZE(consts), NULL) < 0) {
+            return 1;
+        }
+    }
+    return 0;
+}
+
 static inline int
 is_exit_without_lineno(basicblock *b) {
     return b->b_exit && b->b_instr[0].i_lineno < 0;