This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH] Free results of varobj_get_type and type_to_string
- From: Joel Brobecker <brobecker at adacore dot com>
- To: Simon Marchi <simon dot marchi at ericsson dot com>
- Cc: gdb-patches at sourceware dot org
- Date: Thu, 29 Jan 2015 09:11:26 +0400
- Subject: Re: [PATCH] Free results of varobj_get_type and type_to_string
- Authentication-results: sourceware.org; auth=none
- References: <1422383618-8215-1-git-send-email-simon dot marchi at ericsson dot com>
Hi Simon,
On Tue, Jan 27, 2015 at 01:33:38PM -0500, Simon Marchi wrote:
> varobj_get_type and type_to_string return an allocated string, which is
> not freed at a couple of places.
>
> gdb/ChangeLog:
>
> * mi/mi-cmd-var.c (mi_cmd_var_info_type): Free result of
> varobj_get_type.
> (varobj_update_one): Same.
> * varobj.c (update_type_if_necessary): Free curr_type_str and
> new_type_str.
> (varobj_get_type): Specify in comment that the result needs to be
> freed by the caller.
Thanks looking into this. Comments below.
> struct ui_out *uiout = current_uiout;
> struct varobj *var;
> + char *type;
Would you mind renaming this variable "type_name" instead of type.
I don't know about the others, but "type" is now wired into my brain
as to be a "struct type *"...
> @@ -765,7 +769,11 @@ varobj_update_one (struct varobj *var, enum print_values print_values,
> }
>
> if (r->type_changed)
> - ui_out_field_string (uiout, "new_type", varobj_get_type (r->varobj));
> + {
> + char *type = varobj_get_type (r->varobj);
> + ui_out_field_string (uiout, "new_type", type);
> + xfree (type);
> + }
Same here, please.
Also, can you add an empty line after the local declarations?
This is part of GDB's Coding Style.
> /* Obtain the type of an object Variable as a string similar to the one gdb
> - prints on the console. */
> + prints on the console. The caller is responsible for freeing the string.
> + */
Thanks for updating functions' documentation. I really appreciate that.
> char *
> varobj_get_type (struct varobj *var)
> @@ -1303,6 +1304,8 @@ update_type_if_necessary (struct varobj *var, struct value *new_value)
> var->num_children = -1;
> return 1;
> }
> + xfree (curr_type_str);
> + xfree (new_type_str);
In this case, you're still missing the case where the function returns,
I believe.
One way to handle the situation, I think in a way that makes the
allocation + deallocation localized would be to introduce a variable
containing the result of the strcmp? For instance:
new_type_str = type_to_string (new_type);
curr_type_str = varobj_get_type (var);
type_name_has_changed = strcmp (curr_type_str, new_type_str) != 0;
xfree (new_type_str);
xfree (curr_type_str);
if (type_name_has_changed)
{
--
Joel