Fixed bug detected by sp-tests

Cleanup during reviews of new pushed code


BUILD/compile-pentium-debug-max:
  Use --debug=full as default
BUILD/compile-pentium-debug:
  Use --debug=full as default
mysys/my_alloc.c:
  More debugging
sql/item_func.cc:
  Cleanup new code
  Don't call insert_id() for last_insert_id(value) to avoid side effects
sql/item_subselect.cc:
  Fixed DBUG output
sql/sp_head.cc:
  Simple cleanup
sql/sql_lex.cc:
  Moved usage of arguments first in lex_start to make their usage clearer
  Remove sl->expr_list.deleete_elements() becasue:
  - It didn't do anything (delete_elements on a list of list is a no-op operation)
  - The deleted for loop used SELECT_LEX elements that was allocated in mysql_new_select() in sp-head, but freed
  when sphead->mem_root was freed. (delete sphead doesn't remove used SELECT_LEX elements from the global all_selects_list)
sql/sql_parse.cc:
  More DBUG entries

1 files changed, 4 insertions, 2 deletions

diff --git a/mysys/my_alloc.c b/mysys/my_alloc.c
index c14b2899b4b..e0d6288f76b 100644
--- a/mysys/my_alloc.c
+++ b/mysys/my_alloc.c
@@ -166,7 +166,8 @@ gptr alloc_root(MEM_ROOT *mem_root,unsigned int Size)
   gptr point;
   reg1 USED_MEM *next= 0;
   reg2 USED_MEM **prev;
-
+  DBUG_ENTER("alloc_root");
+  DBUG_PRINT("enter",("root: 0x%lx", mem_root));
   DBUG_ASSERT(alloc_root_inited(mem_root));
 
   Size= ALIGN_SIZE(Size);
@@ -213,7 +214,8 @@ gptr alloc_root(MEM_ROOT *mem_root,unsigned int Size)
     mem_root->used= next;
     mem_root->first_block_usage= 0;
   }
-  return(point);
+  DBUG_PRINT("exit",("ptr: 0x%lx", (ulong) point));
+  DBUG_RETURN(point);
 #endif
 }