This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [patch 2/2] Fix watchpoints for multi-inferior
- From: Pedro Alves <alves dot ped at gmail dot com>
- To: Jan Kratochvil <jan dot kratochvil at redhat dot com>
- Cc: gdb-patches at sourceware dot org
- Date: Mon, 02 Jan 2012 19:14:23 +0000
- Subject: Re: [patch 2/2] Fix watchpoints for multi-inferior
- References: <20120102164652.GB10231@host2.jankratochvil.net>
On 01/02/2012 04:46 PM, Jan Kratochvil wrote:
Hi,
this is a new post of:
[patch 4/4] hw watchpoints made multi-inferior
http://sourceware.org/ml/gdb-patches/2010-12/msg00045.html
Currently watchpoints somehow apply to all the inferiors but they do not work
that way anyway. I believe watchpoints should be specific for each inferior.
Yeah.
/* Get data specific for INFERIOR_PTID LWP. Return special data area
for processes being detached. */
static struct i386_inferior_data *
i386_inferior_data_get (void)
{
- /* Intermediate patch stub. */
- static struct i386_inferior_data inf_data_local;
struct inferior *inf = current_inferior ();
- struct i386_inferior_data *inf_data =&inf_data_local;
+ struct i386_inferior_data *inf_data;
+
+ inf_data = inferior_data (inf, i386_inferior_data);
+ if (inf_data == NULL)
+ {
+ inf_data = xzalloc (sizeof (*inf_data));
+ set_inferior_data (current_inferior (), i386_inferior_data, inf_data);
+ }
Do we clear inf_data or inf_data's contents anywhere on inferior
exit or startup, so to not leave debug registers stale across runs?
(The cleanup only runs when the inferior is deleted.)
--- /dev/null
+++ b/gdb/testsuite/gdb.multi/watchpoint-multi.c
@@ -0,0 +1,51 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+ Copyright 2012 Free Software Foundation, Inc.
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 3 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program. If not, see<http://www.gnu.org/licenses/>. */
+
+#include<pthread.h>
+#include<assert.h>
+
+static volatile int a, b, c;
+
+static void
+marker_exit (void)
+{
+ a = 1;
+}
+
+static void *
+start (void *arg)
+{
+ b = 2;
+ c = 3;
+
+ return NULL;
+}
+
+int
+main (void)
+{
+ pthread_t thread;
+ int i;
+
+ i = pthread_create (&thread, NULL, start, NULL);
+ assert (i == 0);
+ i = pthread_join (thread, NULL);
+ assert (i == 0);
+
+ marker_exit ();
+ return 0;
+}
--- /dev/null
+++ b/gdb/testsuite/gdb.multi/watchpoint-multi.exp
@@ -0,0 +1,92 @@
+# Copyright 2012 Free Software Foundation, Inc.
+#
+# This program is free software; you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation; either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program. If not, see<http://www.gnu.org/licenses/>.
+
+if [is_remote target] {
+ # It is KFAIL.
+ continue
Did you mean to turn this into a real kfail? What are the
gdbserver problems, btw?
+
+# Simulate non-stop+target-async which also uses breakpoint always-inserted.
+gdb_test_no_output "set breakpoint always-inserted on"
+# displaced-stepping is also needed as other GDB sometimes still removes the
+# breakpoints, even with always-inserted on.
+gdb_test_no_output "set displaced-stepping on"
'if ![support_displaced_stepping] bail' missing, as well as the usual
hw watchpoints support checks.
Otherwise looks okay.
--
Pedro Alves