This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [MI non-stop 10/11] Skip varobj in running threads.
- From: Pedro Alves <pedro at codesourcery dot com>
- To: gdb-patches at sourceware dot org
- Cc: Vladimir Prus <vladimir at codesourcery dot com>
- Date: Fri, 11 Jul 2008 14:56:39 +0100
- Subject: Re: [MI non-stop 10/11] Skip varobj in running threads.
- References: <200806282100.06214.vladimir@codesourcery.com>
On Saturday 28 June 2008 18:00:06, Vladimir Prus wrote:
>
> const char mi_no_values[] = "--no-values";
> const char mi_simple_values[] = "--simple-values";
> @@ -610,8 +611,19 @@ mi_cmd_var_update (char *command, char **argv, int
> argc) cr = rootlist;
> while (*cr != NULL)
> {
> - if (*name == '*' || varobj_floating_p (*cr))
> - varobj_update_one (*cr, print_values, 0 /* implicit */);
> + int thread_id = varobj_get_thread_id (*cr);
> + int thread_running = 0;
> + if (thread_id == 0 && is_executing (inferior_ptid))
> + thread_running = 1;
> + else if (thread_id > 0)
> + {
> + struct thread_info *tp = find_thread_id (thread_id);
> + if (tp)
> + thread_running = is_running (tp->ptid);
> + }
> + if (!thread_running)
> + if (*name == '*' || varobj_floating_p (*cr))
> + varobj_update_one (*cr, print_values, 0 /* implicit */);
You should reverse the logic, and do the update if the thread is_stopped
instead of !is_running. That is because there is now exited state
to care for as well, and is_running also returns false on those. Essentially,
the is_STATE functions should not be tested with negation, if you're looking
for a particular state.
--
Pedro Alves