add naive unpythonic return type inference to function parsing

examples/execve2.py

@@ -1,11 +1,17 @@
 from pythonbpf.decorators import bpf, section
-from ctypes import c_void_p, c_int32
+from ctypes import c_void_p, c_int64, c_int32
 
 
 @bpf
 @section("tracepoint/syscalls/sys_enter_execve")
 def hello(ctx: c_void_p) -> c_int32:
-    print("Hello, World!")
+    print("entered")
     return c_int32(0)
 
+@bpf
+@section("tracepoint/syscalls/sys_exit_execve")
+def hello_again(ctx: c_void_p) -> c_int64:
+    print("exited")
+    return c_int64(0)
+
 LICENSE = "GPL"

pythonbpf/functions_pass.py

@@ -1,47 +1,6 @@
 from llvmlite import ir
 import ast
-
+from .type_deducer import ctypes_to_ir
-
-def emit_function(module: ir.Module, name: str):
-    ret_type = ir.IntType(32)
-    ptr_type = ir.PointerType()
-    func_ty = ir.FunctionType(ret_type, [ptr_type])
-
-    func = ir.Function(module, func_ty, name)
-
-    param = func.args[0]
-    param.add_attribute("nocapture")
-
-    func.attributes.add("nounwind")
-    # func.attributes.add("\"frame-pointer\"=\"all\"")
-    # func.attributes.add("no-trapping-math", "true")
-    # func.attributes.add("stack-protector-buffer-size", "8")
-
-    block = func.append_basic_block(name="entry")
-    builder = ir.IRBuilder(block)
-    fmt_gvar = module.get_global("hello.____fmt")
-
-    if fmt_gvar is None:
-        # If you haven't created the format string global yet
-        print("Warning: Format string global not found")
-    else:
-        # Cast integer 6 to function pointer type
-        fn_type = ir.FunctionType(ir.IntType(
-            64), [ptr_type, ir.IntType(32)], var_arg=True)
-        fn_ptr_type = ir.PointerType(fn_type)
-        fn_addr = ir.Constant(ir.IntType(64), 6)
-        fn_ptr = builder.inttoptr(fn_addr, fn_ptr_type)
-        # Call the function
-        builder.call(fn_ptr, [fmt_gvar, ir.Constant(ir.IntType(32), 14)])
-
-    builder.ret(ir.Constant(ret_type, 0))
-
-    func.return_value.add_attribute("noundef")
-    func.linkage = "dso_local"
-    func.section = "kprobe/sys_clone"
-    print("function emitted:", name)
-    return func
-
 
 def get_probe_string(func_node):
     """Extract the probe string from the decorator of the function node."""
@@ -58,7 +17,7 @@ def get_probe_string(func_node):
     return "helper"
 
 
-def process_func_body(module, builder, func_node, func):
+def process_func_body(module, builder, func_node, func, ret_type):
     """Process the body of a bpf function"""
     # TODO: A lot.  We just have print -> bpf_trace_printk for now
     did_return = False
@@ -100,23 +59,28 @@ def process_func_body(module, builder, func_node, func):
             if stmt.value is None:
                 builder.ret(ir.Constant(ir.IntType(32), 0))
                 did_return = True
-            elif isinstance(stmt.value, ast.Call) and isinstance(stmt.value.func, ast.Name) and stmt.value.func.id == "c_int32" and len(stmt.value.args) == 1 and isinstance(stmt.value.args[0], ast.Constant) and isinstance(stmt.value.args[0].value, int):
+            elif isinstance(stmt.value, ast.Call) and isinstance(stmt.value.func, ast.Name) and len(stmt.value.args) == 1 and isinstance(stmt.value.args[0], ast.Constant) and isinstance(stmt.value.args[0].value, int):
-                builder.ret(ir.Constant(ir.IntType(
+                call_type = stmt.value.func.id
-                    32), stmt.value.args[0].value))
+                if ctypes_to_ir(call_type) != ret_type:
-                did_return = True
+                    raise ValueError(f"Return type mismatch: expected {ctypes_to_ir(call_type)}, got {call_type}")
+                else:
+                    builder.ret(ir.Constant(ret_type, stmt.value.args[0].value))
+                    did_return = True
             else:
                 print("Unsupported return value")
     if not did_return:
         builder.ret(ir.Constant(ir.IntType(32), 0))
 
 
-def process_bpf_chunk(func_node, module):
+def process_bpf_chunk(func_node, module, return_type):
     """Process a single BPF chunk (function) and emit corresponding LLVM IR."""
 
     func_name = func_node.name
 
-    # TODO: parse return type
+    #TODO: The function actual arg retgurn type is parsed, 
-    ret_type = ir.IntType(32)
+    # but the actual output is not. It's still very wrong. Try uncommenting the 
+    # code in execve2.py once
+    ret_type = return_type
 
     # TODO: parse parameters
     param_types = []
@@ -142,7 +106,7 @@ def process_bpf_chunk(func_node, module):
     block = func.append_basic_block(name="entry")
     builder = ir.IRBuilder(block)
 
-    process_func_body(module, builder, func_node, func)
+    process_func_body(module, builder, func_node, func, ret_type)
 
     print(func)
     print(module)
@@ -154,39 +118,59 @@ def func_proc(tree, module, chunks):
         func_type = get_probe_string(func_node)
         print(f"Found probe_string of {func_node.name}: {func_type}")
 
-        process_bpf_chunk(func_node, module)
+        process_bpf_chunk(func_node, module, ctypes_to_ir(infer_return_type(func_node)))
 
-
+def infer_return_type(func_node: ast.FunctionDef):
-def functions_processing(tree, module):
+    if not isinstance(func_node, (ast.FunctionDef, ast.AsyncFunctionDef)):
-    bpf_functions = []
+        raise TypeError("Expected ast.FunctionDef")
-    helper_functions = []
+    if func_node.returns is not None:
-    for node in tree.body:
+        try:
-        section_name = ""
+            return ast.unparse(func_node.returns)
-        if isinstance(node, ast.FunctionDef):
+        except Exception:
-            if len(node.decorator_list) == 1:
+            node = func_node.returns
-                bpf_functions.append(node)
+            if isinstance(node, ast.Name):
-                node.end_lineno
+                return node.id
-            else:
+            if isinstance(node, ast.Attribute):
-                # IDK why this check is needed, but whatever
+                return getattr(node, "attr", type(node).__name__)
-                if 'helper_functions' not in locals():
+            try:
-                    helper_functions.append(node)
+                return str(node)
-
+            except Exception:
-    # TODO: implement helpers first
+                return type(node).__name__
-
+    found_type = None
-    for func in bpf_functions:
+    def _expr_type(e):
-        dec = func.decorator_list[0]
+        if e is None:
-        if (
+            return "None"
-            isinstance(dec, ast.Call)
+        if isinstance(e, ast.Constant):
-            and isinstance(dec.func, ast.Name)
+            return type(e.value).__name__
-            and dec.func.id == "section"
+        if isinstance(e, ast.Name):
-            and len(dec.args) == 1
+            return e.id
-            and isinstance(dec.args[0], ast.Constant)
+        if isinstance(e, ast.Call):
-            and isinstance(dec.args[0].value, str)
+            f = e.func
-        ):
+            if isinstance(f, ast.Name):
-            section_name = dec.args[0].value
+                return f.id
-        else:
+            if isinstance(f, ast.Attribute):
-            print(f"ERROR: Invalid decorator for function {func.name}")
+                try:
-            continue
+                    return ast.unparse(f)
-
+                except Exception:
-        # TODO: parse arguments and return type
+                    return getattr(f, "attr", type(f).__name__)
-        emit_function(module, func.name + "func")
+            try:
+                return ast.unparse(f)
+            except Exception:
+                return type(f).__name__
+        if isinstance(e, ast.Attribute):
+            try:
+                return ast.unparse(e)
+            except Exception:
+                return getattr(e, "attr", type(e).__name__)
+        try:
+            return ast.unparse(e)
+        except Exception:
+            return type(e).__name__
+    for node in ast.walk(func_node):
+        if isinstance(node, ast.Return):
+            t = _expr_type(node.value)
+            if found_type is None:
+                found_type = t
+            elif found_type != t:
+                raise ValueError(f"Conflicting return types: {found_type} vs {t}")
+    return found_type or "None"

pythonbpf/type_deducer.py

@@ -1,29 +1,21 @@
-import ctypes
 from llvmlite import ir
 
-def ctypes_to_ir(ctype):
+#TODO: THIS IS NOT SUPPOSED TO MATCH STRINGS :skull:
-    if ctype is ctypes.c_int32:
+def ctypes_to_ir(ctype: str):
-        return ir.IntType(32)
+    print("CTYPE", ctype)
-    if ctype is ctypes.c_int64:
+    mapping = {
-        return ir.IntType(64)
+        "c_int8": ir.IntType(8),
-    if ctype is ctypes.c_uint8:
+        "c_uint8": ir.IntType(8),
-        return ir.IntType(8)
+        "c_int16": ir.IntType(16),
-    if ctype is ctypes.c_double:
+        "c_uint16": ir.IntType(16),
-        return ir.DoubleType()
+        "c_int32": ir.IntType(32),
-    if ctype is ctypes.c_float:
+        "c_uint32": ir.IntType(32),
-        return ir.FloatType()
+        "c_int64": ir.IntType(64),
-
+        "c_uint64": ir.IntType(64),
-    # pointers
+        "c_float": ir.FloatType(),
-    if hasattr(ctype, "_type_") and hasattr(ctype, "_length_"):
+        "c_double": ir.DoubleType(),
-        # ctypes array
+        "c_void_p": ir.IntType(64),
-        return ir.ArrayType(ctypes_to_ir(ctype._type_), ctype._length_)
+    }
-
+    if ctype in mapping:
-    # if hasattr(ctype, "_type_") and issubclass(ctype, ctypes._Pointer):
+        return mapping[ctype]
-    #     return ir.PointerType(ctypes_to_ir(ctype._type_))
-
-    # structs
-    if issubclass(ctype, ctypes.Structure):
-        fields = [ctypes_to_ir(f[1]) for f in ctype._fields_]
-        return ir.LiteralStructType(fields)
-
     raise NotImplementedError(f"No mapping for {ctype}")