Use correct size for stack slot in inline asm codegen

Author: bjorn3

src/inline_asm.rs

@@ -871,7 +871,8 @@ fn call_inline_asm<'tcx>(
     inputs: Vec<(Size, Value)>,
     outputs: Vec<(Size, CPlace<'tcx>)>,
 ) {
-    let stack_slot = fx.create_stack_slot(u32::try_from(slot_size.bytes()).unwrap(), 16);
+    let stack_slot =
+        fx.create_stack_slot(u32::try_from(slot_size.bytes().next_multiple_of(16)).unwrap(), 16);
 
     let inline_asm_func = fx
         .module