gh-98831: rewrite pattern matching opcodes in the instruction definition DSL (#101287)

Author: iritkatriel

Python/bytecodes.c

@@ -2092,61 +2092,37 @@ dummy_func(
             PUSH(len_o);
         }
 
-        // stack effect: (__0, __1 -- )
-        inst(MATCH_CLASS) {
+        inst(MATCH_CLASS, (subject, type, names -- attrs)) {
             // Pop TOS and TOS1. Set TOS to a tuple of attributes on success, or
             // None on failure.
-            PyObject *names = POP();
-            PyObject *type = POP();
-            PyObject *subject = TOP();
             assert(PyTuple_CheckExact(names));
-            PyObject *attrs = match_class(tstate, subject, type, oparg, names);
-            Py_DECREF(names);
-            Py_DECREF(type);
+            attrs = match_class(tstate, subject, type, oparg, names);
+            DECREF_INPUTS();
             if (attrs) {
-                // Success!
-                assert(PyTuple_CheckExact(attrs));
-                SET_TOP(attrs);
-            }
-            else if (_PyErr_Occurred(tstate)) {
-                // Error!
-                goto error;
+                assert(PyTuple_CheckExact(attrs));  // Success!
             }
             else {
-                // Failure!
-                SET_TOP(Py_NewRef(Py_None));
+                ERROR_IF(_PyErr_Occurred(tstate), error);  // Error!
+                attrs = Py_NewRef(Py_None);  // Failure!
             }
-            Py_DECREF(subject);
         }
 
-        // stack effect: ( -- __0)
-        inst(MATCH_MAPPING) {
-            PyObject *subject = TOP();
+        inst(MATCH_MAPPING, (subject -- subject, res)) {
             int match = Py_TYPE(subject)->tp_flags & Py_TPFLAGS_MAPPING;
-            PyObject *res = match ? Py_True : Py_False;
-            PUSH(Py_NewRef(res));
+            res = Py_NewRef(match ? Py_True : Py_False);
             PREDICT(POP_JUMP_IF_FALSE);
         }
 
-        // stack effect: ( -- __0)
-        inst(MATCH_SEQUENCE) {
-            PyObject *subject = TOP();
+        inst(MATCH_SEQUENCE, (subject -- subject, res)) {
             int match = Py_TYPE(subject)->tp_flags & Py_TPFLAGS_SEQUENCE;
-            PyObject *res = match ? Py_True : Py_False;
-            PUSH(Py_NewRef(res));
+            res = Py_NewRef(match ? Py_True : Py_False);
             PREDICT(POP_JUMP_IF_FALSE);
         }
 
-        // stack effect: ( -- __0)
-        inst(MATCH_KEYS) {
+        inst(MATCH_KEYS, (subject, keys -- subject, keys, values_or_none)) {
             // On successful match, PUSH(values). Otherwise, PUSH(None).
-            PyObject *keys = TOP();
-            PyObject *subject = SECOND();
-            PyObject *values_or_none = match_keys(tstate, subject, keys);
-            if (values_or_none == NULL) {
-                goto error;
-            }
-            PUSH(values_or_none);
+            values_or_none = match_keys(tstate, subject, keys);
+            ERROR_IF(values_or_none == NULL, error);
         }
 
         // stack effect: ( -- )

Python/generated_cases.c.h

@@ -133,10 +133,10 @@ static const struct {
     [JUMP_IF_TRUE_OR_POP] = { -1, -1, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IB },
     [JUMP_BACKWARD_NO_INTERRUPT] = { -1, -1, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IB },
     [GET_LEN] = { -1, -1, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IX },
-    [MATCH_CLASS] = { -1, -1, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IB },
-    [MATCH_MAPPING] = { -1, -1, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IX },
-    [MATCH_SEQUENCE] = { -1, -1, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IX },
-    [MATCH_KEYS] = { -1, -1, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IX },
+    [MATCH_CLASS] = { 3, 1, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IB },
+    [MATCH_MAPPING] = { 1, 2, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IX },
+    [MATCH_SEQUENCE] = { 1, 2, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IX },
+    [MATCH_KEYS] = { 2, 3, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IX },
     [GET_ITER] = { -1, -1, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IX },
     [GET_YIELD_FROM_ITER] = { -1, -1, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IX },
     [FOR_ITER] = { -1, -1, DIR_NONE, DIR_NONE, DIR_NONE, true, INSTR_FMT_IB },

Python/opcode_metadata.h

@@ -26,7 +26,7 @@
 )
 BEGIN_MARKER = "// BEGIN BYTECODES //"
 END_MARKER = "// END BYTECODES //"
-RE_PREDICTED = r"^\s*(?:PREDICT\(|GO_TO_INSTRUCTION\(|DEOPT_IF\(.*?,\s*)(\w+)\);\s*$"
+RE_PREDICTED = r"^\s*(?:PREDICT\(|GO_TO_INSTRUCTION\(|DEOPT_IF\(.*?,\s*)(\w+)\);\s*(?://.*)?$"
 UNUSED = "unused"
 BITS_PER_CODE_UNIT = 16
 
@@ -354,7 +354,7 @@ def write_body(self, out: Formatter, dedent: int, cache_adjust: int = 0) -> None
         assert dedent <= 0
         extra = " " * -dedent
         for line in self.block_text:
-            if m := re.match(r"(\s*)ERROR_IF\((.+), (\w+)\);\s*$", line):
+            if m := re.match(r"(\s*)ERROR_IF\((.+), (\w+)\);\s*(?://.*)?$", line):
                 space, cond, label = m.groups()
                 # ERROR_IF() must pop the inputs from the stack.
                 # The code block is responsible for DECREF()ing them.
@@ -378,7 +378,7 @@ def write_body(self, out: Formatter, dedent: int, cache_adjust: int = 0) -> None
                     )
                 else:
                     out.write_raw(f"{extra}{space}if ({cond}) goto {label};\n")
-            elif m := re.match(r"(\s*)DECREF_INPUTS\(\);\s*$", line):
+            elif m := re.match(r"(\s*)DECREF_INPUTS\(\);\s*(?://.*)?$", line):
                 if not self.register:
                     space = m.group(1)
                     for ieff in self.input_effects:
@@ -964,7 +964,7 @@ def extract_block_text(block: parser.Block) -> tuple[list[str], list[str]]:
 
     # Separate PREDICT(...) macros from end
     predictions: list[str] = []
-    while blocklines and (m := re.match(r"^\s*PREDICT\((\w+)\);\s*$", blocklines[-1])):
+    while blocklines and (m := re.match(r"^\s*PREDICT\((\w+)\);\s*(?://.*)?$", blocklines[-1])):
         predictions.insert(0, m.group(1))
         blocklines.pop()
 

Tools/cases_generator/generate_cases.py

@@ -215,6 +215,20 @@ def test_error_if_plain():
     """
     run_cases_test(input, output)
 
+def test_error_if_plain_with_comment():
+    input = """
+        inst(OP, (--)) {
+            ERROR_IF(cond, label);  // Comment is ok
+        }
+    """
+    output = """
+        TARGET(OP) {
+            if (cond) goto label;
+            DISPATCH();
+        }
+    """
+    run_cases_test(input, output)
+
 def test_error_if_pop():
     input = """
         inst(OP, (left, right -- res)) {