Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
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"
+
+   


  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