From: "Pierre Muller" <muller@ics.u-strasbg.fr>
To: "'Eli Zaretskii'" <eliz@gnu.org>,
"'Joel Brobecker'" <brobecker@adacore.com>,
"'Mark Kettenis'" <mark.kettenis@xs4all.nl>
Cc: <gdb-patches@sourceware.org>
Subject: [RFA-v3] Fix a windows bug if two watchpoints are used
Date: Tue, 22 Sep 2009 12:55:00 -0000 [thread overview]
Message-ID: <000301ca3b83$d15fee00$741fca00$@u-strasbg.fr> (raw)
In-Reply-To: <E1MIuRD-0001QO-Dj@fencepost.gnu.org>
Mark never answered that thread,
as a result, this one line patch that fixes
a problem with multiple watches for windows
is not in 7.0 branch :(
The thread started here:
http://sourceware.org/ml/gdb-patches/2009-06/msg00054.html
with
http://sourceware.org/ml/gdb-patches/2009-06/msg00416.html
containing the final one code line patch.
Mark, could you give your impression?
Thanks in advance,
Pierre Muller
Pascal language support maintainer for GDB
PS: I tested the patch on Compile Farm, on a amd64 linux,
no failure on new watchpoints.exp test,
nor on CVS head nor with patched version.
2009-09-22 Pierre Muller <muller@ics.u-strasbg.fr>
* i386-nat.c (i386_stopped_data_address): Also check that
hitted watch register is not vacant.
gdb/testsuite ChangeLog entry:
2009-09-22 Pierre Muller <muller@ics.u-strasbg.fr>
New test for two watchpoints, with disabling of
the first inserted.
* testsuite/gdb.base/watchpoints.c: New file.
* testsuite/gdb.base/watchpoints.exp: New file.
Index: src/gdb/i386-nat.c
===================================================================
RCS file: /cvs/src/src/gdb/i386-nat.c,v
retrieving revision 1.24
diff -u -p -r1.24 i386-nat.c
--- src/gdb/i386-nat.c 2 Jul 2009 17:21:06 -0000 1.24
+++ src/gdb/i386-nat.c 22 Sep 2009 08:25:36 -0000
@@ -555,7 +555,10 @@ i386_stopped_data_address (struct target
that GDB doesn't call the target_stopped_data_address
method except for data watchpoints. In other words, I'm
being paranoiac. */
- && I386_DR_GET_RW_LEN (i) != 0)
+ && I386_DR_GET_RW_LEN (i) != 0
+ /* This third condition makes sure DRi is not vacant, this
+ avoids false positives in windows-nat.c. */
+ && !I386_DR_VACANT (i))
{
addr = dr_mirror[i];
rc = 1;
Index: src/gdb/testsuite/gdb.base/watchpoints.c
===================================================================
RCS file: testsuite/gdb.base/watchpoints.c
diff -N testsuite/gdb.base/watchpoints.c
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ src/gdb/testsuite/gdb.base/watchpoints.c 25 Jun 2009 22:44:17 -0000
@@ -0,0 +1,33 @@
+#include <stdio.h>
+#include <unistd.h>
+/*
+ * This source is mainly to test
+ * what happens when a watchpoint is removed
+ * while another watchpoint, inserted later is
+ * left active.
+ */
+
+int count = -1;
+int ival1 = -1;
+int ival2 = -1;
+int ival3 = -1;
+int ival4 = -1;
+
+int
+main ()
+{
+#ifdef usestubs
+ set_debug_traps();
+ breakpoint();
+#endif
+
+ for (count = 0; count < 4; count++) {
+ ival1 = count; ival2 = count;
+ ival3 = count; ival4 = count;
+ }
+
+ ival1 = count; ival2 = count; /* Outside loop */
+ ival3 = count; ival4 = count;
+
+ return 0;
+}
Index: src/gdb/testsuite/gdb.base/watchpoints.exp
===================================================================
RCS file: testsuite/gdb.base/watchpoints.exp
diff -N testsuite/gdb.base/watchpoints.exp
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ src/gdb/testsuite/gdb.base/watchpoints.exp 25 Jun 2009 22:44:17 -0000
@@ -0,0 +1,101 @@
+# Copyright 2009
+# 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/>.
+
+# This file was written by Pierre Muller. (muller@ics.u-strasbg.fr)
+
+if $tracelevel then {
+ strace $tracelevel
+}
+
+set prms_id 0
+set bug_id 0
+
+set testfile "watchpoints"
+set srcfile ${testfile}.c
+set binfile ${objdir}/${subdir}/${testfile}
+
+set wp_set 1
+
+if [get_compiler_info ${binfile}] {
+ return -1
+}
+
+if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable
{debug}] != "" } {
+ untested watchpoint.exp
+ return -1
+}
+
+# Start with a fresh gdb.
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load $binfile
+
+ runto_main
+ gdb_test "watch ival1" "" ""
+ gdb_test "watch ival3" "" ""
+
+ set timeout 600
+
+ gdb_test "cont" "Continuing.*\[Ww\]atchpoint.*ival1.*Old value =
-1.*New value = 0.*ival1 = count; ival2 = count;.*" "watchpoint hit, first
time"
+
+ gdb_test "cont" "Continuing.*\[Ww\]atchpoint.*ival3.*Old value =
-1.*New value = 0.*ival3 = count; ival4 = count;.*" "watchpoint hit, first
time"
+
+ # Check that the ival3 hit count is reported correctly
+ gdb_test "info break" ".*watchpoint\[ \t\]+keep\[ \t\]+y\[
\t\]+ival3\r\n\[ \t]+breakpoint already hit 1 time.*" "Watchpoint hit count
is 1"
+
+ # Continue until the next change for ival1, from 0 to 1.
+ gdb_test "cont" "Continuing.*\[Ww\]atchpoint.*ival1.*Old value = 0.*New
value = 1.*ival1 = count; ival2 = count;.*" "watchpoint ival1 hit, second
time"
+
+ # Check that the hit count for ival1 is reported correctly
+ gdb_test "info break" ".*watchpoint\[ \t\]+keep\[ \t\]+y\[
\t\]+ival1\r\n\[ \t]+breakpoint already hit 2 times.*" "Watchpoint ival1 hit
count is 2"
+
+ # Continue until the next change for ival3, from 0 to 1.
+ gdb_test "cont" "Continuing.*\[Ww\]atchpoint.*ival3.*Old value = 0.*New
value = 1.*ival3 = count; ival4 = count;.*" "watchpoint hit, second time"
+
+ # Check that the hit count is reported correctly
+ gdb_test "info break" ".*watchpoint\[ \t\]+keep\[ \t\]+y\[
\t\]+ival3\r\n\[ \t]+breakpoint already hit 2 times.*" "Watchpoint hit count
is 2"
+
+ # Continue until the next change, from 1 to 2.
+ gdb_test "cont" "Continuing.*\[Ww\]atchpoint.*ival1.*Old value = 1.*New
value = 2.*ival1 = count; ival2 = count;.*" "watchpoint ival1 hit, third
time"
+
+ # Check that the hit count is reported correctly
+ gdb_test "info break" ".*watchpoint\[ \t\]+keep\[ \t\]+y\[
\t\]+ival1\r\n\[ \t]+breakpoint already hit 3 times.*" "Watchpoint ival1 hit
count is 3"
+ # Disable ival1 watchpoint
+ gdb_test "disable 2" "" ""
+
+ # Continue until the next change, from 1 to 2.
+ gdb_test "cont" "Continuing.*\[Ww\]atchpoint.*ival3.*Old value = 1.*New
value = 2.*ival3 = count; ival4 = count;.*" "watchpoint hit, third time"
+
+ # Check that the hit count is reported correctly
+ gdb_test "info break" ".*watchpoint\[ \t\]+keep\[ \t\]+y\[
\t\]+ival3\r\n\[ \t]+breakpoint already hit 3 times.*" "Watchpoint hit count
is 3"
+
+ # Continue until the next change, from 2 to 3.
+ gdb_test "cont" "Continuing.*\[Ww\]atchpoint.*ival3.*Old value = 2.*New
value = 3.*ival3 = count; ival4 = count;.*" "watchpoint hit, fourth time"
+
+ # Check that the hit count is reported correctly
+ gdb_test "info break" ".*watchpoint\[ \t\]+keep\[ \t\]+y\[
\t\]+ival3\r\n\[ \t]+breakpoint already hit 4 times.*" "Watchpoint hit count
is 4"
+
+ # Continue until the next change, from 3 to 4.
+ # Note that this one is outside the loop.
+
+ gdb_test "cont" "Continuing.*\[Ww\]atchpoint.*ival3.*Old value = 3.*New
value = 4.*ival3 = count; ival4 = count;.*" "watchpoint hit, fifth time"
+
+ # Check that the hit count is reported correctly
+ gdb_test "info break" ".*watchpoint\[ \t\]+keep\[ \t\]+y\[
\t\]+ival3\r\n\[ \t]+breakpoint already hit 5 times.*" "Watchpoint hit count
is 5"
+
+
next prev parent reply other threads:[~2009-09-22 12:55 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-06-03 22:58 [RFA] " Pierre Muller
2009-06-03 23:21 ` Eli Zaretskii
2009-06-03 23:33 ` Pierre Muller
2009-06-04 3:16 ` Eli Zaretskii
2009-06-04 3:29 ` Christopher Faylor
2009-06-04 6:34 ` [RFA-v2] " Pierre Muller
2009-06-04 7:12 ` Eli Zaretskii
2009-06-04 7:33 ` Pierre Muller
2009-06-04 10:31 ` Eli Zaretskii
2009-06-04 14:52 ` Pierre Muller
2009-06-04 15:22 ` Eli Zaretskii
2009-06-16 22:43 ` [PING][RFA-v2] " Pierre Muller
2009-06-22 20:56 ` Joel Brobecker
2009-06-23 1:04 ` Eli Zaretskii
2009-09-22 12:55 ` Pierre Muller [this message]
2009-09-24 3:04 ` [RFA-v3] " Eli Zaretskii
2009-09-24 17:54 ` Joel Brobecker
2009-09-24 22:08 ` Pierre Muller
2009-09-25 1:53 ` Joel Brobecker
2009-09-25 15:32 ` Pierre Muller
2009-09-25 16:07 ` Joel Brobecker
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='000301ca3b83$d15fee00$741fca00$@u-strasbg.fr' \
--to=muller@ics.u-strasbg.fr \
--cc=brobecker@adacore.com \
--cc=eliz@gnu.org \
--cc=gdb-patches@sourceware.org \
--cc=mark.kettenis@xs4all.nl \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox