GH-118095: Handle `RETURN_GENERATOR` in tier 2 (GH-118180)

9 files changed, 111 insertions, 72 deletions

diff --git a/Python/bytecodes.c b/Python/bytecodes.c
index c31617d..4855049 100644
--- a/Python/bytecodes.c
+++ b/Python/bytecodes.c
@@ -837,12 +837,7 @@ dummy_func(
             _PyFrame_StackPush(frame, retval);
             LOAD_SP();
             LOAD_IP(frame->return_offset);
-#if LLTRACE && TIER_ONE
-            lltrace = maybe_lltrace_resume_frame(frame, &entry_frame, GLOBALS());
-            if (lltrace < 0) {
-                goto exit_unwind;
-            }
-#endif
+            LLTRACE_RESUME_FRAME();
         }
 
         macro(RETURN_VALUE) =
@@ -3186,12 +3181,7 @@ dummy_func(
             tstate->py_recursion_remaining--;
             LOAD_SP();
             LOAD_IP(0);
-#if LLTRACE && TIER_ONE
-            lltrace = maybe_lltrace_resume_frame(frame, &entry_frame, GLOBALS());
-            if (lltrace < 0) {
-                goto exit_unwind;
-            }
-#endif
+            LLTRACE_RESUME_FRAME();
         }
 
         macro(CALL_BOUND_METHOD_EXACT_ARGS) =
@@ -3877,7 +3867,7 @@ dummy_func(
             }
         }
 
-        tier1 inst(RETURN_GENERATOR, (--)) {
+        inst(RETURN_GENERATOR, (-- res)) {
             assert(PyFunction_Check(frame->f_funcobj));
             PyFunctionObject *func = (PyFunctionObject *)frame->f_funcobj;
             PyGenObject *gen = (PyGenObject *)_Py_MakeCoro(func);
@@ -3887,19 +3877,19 @@ dummy_func(
             assert(EMPTY());
             _PyFrame_SetStackPointer(frame, stack_pointer);
             _PyInterpreterFrame *gen_frame = (_PyInterpreterFrame *)gen->gi_iframe;
-            frame->instr_ptr = next_instr;
+            frame->instr_ptr++;
             _PyFrame_Copy(frame, gen_frame);
             assert(frame->frame_obj == NULL);
             gen->gi_frame_state = FRAME_CREATED;
             gen_frame->owner = FRAME_OWNED_BY_GENERATOR;
             _Py_LeaveRecursiveCallPy(tstate);
-            assert(frame != &entry_frame);
+            res = (PyObject *)gen;
             _PyInterpreterFrame *prev = frame->previous;
             _PyThreadState_PopFrame(tstate, frame);
             frame = tstate->current_frame = prev;
-            _PyFrame_StackPush(frame, (PyObject *)gen);
             LOAD_IP(frame->return_offset);
-            goto resume_frame;
+            LOAD_SP();
+            LLTRACE_RESUME_FRAME();
         }
 
         inst(BUILD_SLICE, (start, stop, step if (oparg == 3) -- slice)) {
diff --git a/Python/ceval_macros.h b/Python/ceval_macros.h
index 224cd1d..871d174 100644
--- a/Python/ceval_macros.h
+++ b/Python/ceval_macros.h
@@ -86,6 +86,18 @@
 #define PRE_DISPATCH_GOTO() ((void)0)
 #endif
 
+#if LLTRACE
+#define LLTRACE_RESUME_FRAME() \
+do { \
+    lltrace = maybe_lltrace_resume_frame(frame, &entry_frame, GLOBALS()); \
+    if (lltrace < 0) { \
+        goto exit_unwind; \
+    } \
+} while (0)
+#else
+#define LLTRACE_RESUME_FRAME() ((void)0)
+#endif
+
 #ifdef Py_GIL_DISABLED
 #define QSBR_QUIESCENT_STATE(tstate) _Py_qsbr_quiescent_state(((_PyThreadStateImpl *)tstate)->qsbr)
 #else
diff --git a/Python/executor_cases.c.h b/Python/executor_cases.c.h
index 7403d6f..1eb3da9 100644
--- a/Python/executor_cases.c.h
+++ b/Python/executor_cases.c.h
@@ -988,12 +988,7 @@
             _PyFrame_StackPush(frame, retval);
             LOAD_SP();
             LOAD_IP(frame->return_offset);
-            #if LLTRACE && TIER_ONE
-            lltrace = maybe_lltrace_resume_frame(frame, &entry_frame, GLOBALS());
-            if (lltrace < 0) {
-                goto exit_unwind;
-            }
-            #endif
+            LLTRACE_RESUME_FRAME();
             break;
         }
 
@@ -3213,12 +3208,7 @@
             tstate->py_recursion_remaining--;
             LOAD_SP();
             LOAD_IP(0);
-            #if LLTRACE && TIER_ONE
-            lltrace = maybe_lltrace_resume_frame(frame, &entry_frame, GLOBALS());
-            if (lltrace < 0) {
-                goto exit_unwind;
-            }
-            #endif
+            LLTRACE_RESUME_FRAME();
             break;
         }
 
@@ -3833,6 +3823,35 @@
             break;
         }
 
+        case _RETURN_GENERATOR: {
+            PyObject *res;
+            assert(PyFunction_Check(frame->f_funcobj));
+            PyFunctionObject *func = (PyFunctionObject *)frame->f_funcobj;
+            PyGenObject *gen = (PyGenObject *)_Py_MakeCoro(func);
+            if (gen == NULL) {
+                JUMP_TO_ERROR();
+            }
+            assert(EMPTY());
+            _PyFrame_SetStackPointer(frame, stack_pointer);
+            _PyInterpreterFrame *gen_frame = (_PyInterpreterFrame *)gen->gi_iframe;
+            frame->instr_ptr++;
+            _PyFrame_Copy(frame, gen_frame);
+            assert(frame->frame_obj == NULL);
+            gen->gi_frame_state = FRAME_CREATED;
+            gen_frame->owner = FRAME_OWNED_BY_GENERATOR;
+            _Py_LeaveRecursiveCallPy(tstate);
+            res = (PyObject *)gen;
+            _PyInterpreterFrame *prev = frame->previous;
+            _PyThreadState_PopFrame(tstate, frame);
+            frame = tstate->current_frame = prev;
+            LOAD_IP(frame->return_offset);
+            LOAD_SP();
+            LLTRACE_RESUME_FRAME();
+            stack_pointer[0] = res;
+            stack_pointer += 1;
+            break;
+        }
+
         case _BUILD_SLICE: {
             PyObject *step = NULL;
             PyObject *stop;
diff --git a/Python/frame.c b/Python/frame.c
index f88a8f0..db9d133 100644
--- a/Python/frame.c
+++ b/Python/frame.c
@@ -53,18 +53,6 @@ _PyFrame_MakeAndSetFrameObject(_PyInterpreterFrame *frame)
     return f;
 }
 
-void
-_PyFrame_Copy(_PyInterpreterFrame *src, _PyInterpreterFrame *dest)
-{
-    assert(src->stacktop >= _PyFrame_GetCode(src)->co_nlocalsplus);
-    Py_ssize_t size = ((char*)&src->localsplus[src->stacktop]) - (char *)src;
-    memcpy(dest, src, size);
-    // Don't leave a dangling pointer to the old frame when creating generators
-    // and coroutines:
-    dest->previous = NULL;
-}
-
-
 static void
 take_ownership(PyFrameObject *f, _PyInterpreterFrame *frame)
 {
diff --git a/Python/generated_cases.c.h b/Python/generated_cases.c.h
index 058cac8..0c58f3f 100644
--- a/Python/generated_cases.c.h
+++ b/Python/generated_cases.c.h
@@ -997,12 +997,7 @@
                 tstate->py_recursion_remaining--;
                 LOAD_SP();
                 LOAD_IP(0);
-                #if LLTRACE && TIER_ONE
-                lltrace = maybe_lltrace_resume_frame(frame, &entry_frame, GLOBALS());
-                if (lltrace < 0) {
-                    goto exit_unwind;
-                }
-                #endif
+                LLTRACE_RESUME_FRAME();
             }
             DISPATCH();
         }
@@ -1786,12 +1781,7 @@
                 tstate->py_recursion_remaining--;
                 LOAD_SP();
                 LOAD_IP(0);
-                #if LLTRACE && TIER_ONE
-                lltrace = maybe_lltrace_resume_frame(frame, &entry_frame, GLOBALS());
-                if (lltrace < 0) {
-                    goto exit_unwind;
-                }
-                #endif
+                LLTRACE_RESUME_FRAME();
             }
             DISPATCH();
         }
@@ -4992,12 +4982,7 @@
                 _PyFrame_StackPush(frame, retval);
                 LOAD_SP();
                 LOAD_IP(frame->return_offset);
-                #if LLTRACE && TIER_ONE
-                lltrace = maybe_lltrace_resume_frame(frame, &entry_frame, GLOBALS());
-                if (lltrace < 0) {
-                    goto exit_unwind;
-                }
-                #endif
+                LLTRACE_RESUME_FRAME();
             }
             DISPATCH();
         }
@@ -5006,6 +4991,7 @@
             frame->instr_ptr = next_instr;
             next_instr += 1;
             INSTRUCTION_STATS(RETURN_GENERATOR);
+            PyObject *res;
             assert(PyFunction_Check(frame->f_funcobj));
             PyFunctionObject *func = (PyFunctionObject *)frame->f_funcobj;
             PyGenObject *gen = (PyGenObject *)_Py_MakeCoro(func);
@@ -5015,19 +5001,22 @@
             assert(EMPTY());
             _PyFrame_SetStackPointer(frame, stack_pointer);
             _PyInterpreterFrame *gen_frame = (_PyInterpreterFrame *)gen->gi_iframe;
-            frame->instr_ptr = next_instr;
+            frame->instr_ptr++;
             _PyFrame_Copy(frame, gen_frame);
             assert(frame->frame_obj == NULL);
             gen->gi_frame_state = FRAME_CREATED;
             gen_frame->owner = FRAME_OWNED_BY_GENERATOR;
             _Py_LeaveRecursiveCallPy(tstate);
-            assert(frame != &entry_frame);
+            res = (PyObject *)gen;
             _PyInterpreterFrame *prev = frame->previous;
             _PyThreadState_PopFrame(tstate, frame);
             frame = tstate->current_frame = prev;
-            _PyFrame_StackPush(frame, (PyObject *)gen);
             LOAD_IP(frame->return_offset);
-            goto resume_frame;
+            LOAD_SP();
+            LLTRACE_RESUME_FRAME();
+            stack_pointer[0] = res;
+            stack_pointer += 1;
+            DISPATCH();
         }
 
         TARGET(RETURN_VALUE) {
@@ -5050,12 +5039,7 @@
             _PyFrame_StackPush(frame, retval);
             LOAD_SP();
             LOAD_IP(frame->return_offset);
-            #if LLTRACE && TIER_ONE
-            lltrace = maybe_lltrace_resume_frame(frame, &entry_frame, GLOBALS());
-            if (lltrace < 0) {
-                goto exit_unwind;
-            }
-            #endif
+            LLTRACE_RESUME_FRAME();
             DISPATCH();
         }
 
diff --git a/Python/optimizer.c b/Python/optimizer.c
index b17c299..e5c70f7 100644
--- a/Python/optimizer.c
+++ b/Python/optimizer.c
@@ -697,7 +697,8 @@ top:  // Jump here after _PUSH_FRAME or likely branches
                     // Reserve space for nuops (+ _SET_IP + _EXIT_TRACE)
                     int nuops = expansion->nuops;
                     RESERVE(nuops + 1); /* One extra for exit */
-                    if (expansion->uops[nuops-1].uop == _POP_FRAME) {
+                    int16_t last_op = expansion->uops[nuops-1].uop;
+                    if (last_op == _POP_FRAME || last_op == _RETURN_GENERATOR) {
                         // Check for trace stack underflow now:
                         // We can't bail e.g. in the middle of
                         // LOAD_CONST + _POP_FRAME.
@@ -756,7 +757,7 @@ top:  // Jump here after _PUSH_FRAME or likely branches
                                 Py_FatalError("garbled expansion");
                         }
 
-                        if (uop == _POP_FRAME) {
+                        if (uop == _POP_FRAME || uop == _RETURN_GENERATOR) {
                             TRACE_STACK_POP();
                             /* Set the operand to the function or code object returned to,
                              * to assist optimization passes. (See _PUSH_FRAME below.)
diff --git a/Python/optimizer_analysis.c b/Python/optimizer_analysis.c
index a76edd6..9315d72 100644
--- a/Python/optimizer_analysis.c
+++ b/Python/optimizer_analysis.c
@@ -369,7 +369,7 @@ eliminate_pop_guard(_PyUOpInstruction *this_instr, bool exit)
 static PyCodeObject *
 get_code(_PyUOpInstruction *op)
 {
-    assert(op->opcode == _PUSH_FRAME || op->opcode == _POP_FRAME);
+    assert(op->opcode == _PUSH_FRAME || op->opcode == _POP_FRAME || op->opcode == _RETURN_GENERATOR);
     PyCodeObject *co = NULL;
     uint64_t operand = op->operand;
     if (operand == 0) {
diff --git a/Python/optimizer_bytecodes.c b/Python/optimizer_bytecodes.c
index 481fb83..8bc5634 100644
--- a/Python/optimizer_bytecodes.c
+++ b/Python/optimizer_bytecodes.c
@@ -651,6 +651,28 @@ dummy_func(void) {
         }
     }
 
+    op(_RETURN_GENERATOR, ( -- res)) {
+        SYNC_SP();
+        ctx->frame->stack_pointer = stack_pointer;
+        frame_pop(ctx);
+        stack_pointer = ctx->frame->stack_pointer;
+        OUT_OF_SPACE_IF_NULL(res = sym_new_unknown(ctx));
+
+        /* Stack space handling */
+        assert(corresponding_check_stack == NULL);
+        assert(co != NULL);
+        int framesize = co->co_framesize;
+        assert(framesize > 0);
+        assert(framesize <= curr_space);
+        curr_space -= framesize;
+
+        co = get_code(this_instr);
+        if (co == NULL) {
+            // might be impossible, but bailing is still safe
+            goto done;
+        }
+    }
+
     op(_CHECK_STACK_SPACE, ( --)) {
         assert(corresponding_check_stack == NULL);
         corresponding_check_stack = this_instr;
diff --git a/Python/optimizer_cases.c.h b/Python/optimizer_cases.c.h
index 0a7d96d..4f0941a 100644
--- a/Python/optimizer_cases.c.h
+++ b/Python/optimizer_cases.c.h
@@ -1840,6 +1840,29 @@
             break;
         }
 
+        case _RETURN_GENERATOR: {
+            _Py_UopsSymbol *res;
+            ctx->frame->stack_pointer = stack_pointer;
+            frame_pop(ctx);
+            stack_pointer = ctx->frame->stack_pointer;
+            OUT_OF_SPACE_IF_NULL(res = sym_new_unknown(ctx));
+            /* Stack space handling */
+            assert(corresponding_check_stack == NULL);
+            assert(co != NULL);
+            int framesize = co->co_framesize;
+            assert(framesize > 0);
+            assert(framesize <= curr_space);
+            curr_space -= framesize;
+            co = get_code(this_instr);
+            if (co == NULL) {
+                // might be impossible, but bailing is still safe
+                goto done;
+            }
+            stack_pointer[0] = res;
+            stack_pointer += 1;
+            break;
+        }
+
         case _BUILD_SLICE: {
             _Py_UopsSymbol *slice;
             slice = sym_new_not_null(ctx);