Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: dje@google.com (Doug Evans)
To: gdb-patches@sourceware.org
Subject: [RFC] add h/w breakpoints to x86-linux gdbserver
Date: Tue, 30 Jun 2009 10:54:00 -0000	[thread overview]
Message-ID: <20090630105351.5FBD7846CA@localhost> (raw)

Hi.

This patch is just an experiment in adding h/w breakpoints
to x86-linux gdbserver.

It requires this patch:
http://sourceware.org/ml/gdb-patches/2009-06/msg00859.html

One open issue is that the "exec" target doesn't support h/w breakpoints
so "hbreak foo" fails when the program isn't running, but if you
get the program running first, then hbreak works.

2009-06-30  Doug Evans  <dje@google.com>

	Add h/w breakpoints to x86-linux gdbserver.
	* i386-low.h (i386_low_insert_hw_breakpoint): Declare.
	(i386_low_remove_hw_breakpoint): Declare.
	* i386-low.c (i386_low_insert_hw_breakpoint): New function.
	(i386_low_remove_hw_breakpoint): New function.
	* linux-x86-low.c (x86_insert_point): Handle h/w breakpoint packets.
	(x86_remove_point): Ditto.

	* gdb.base/hbreak.exp: New file.
	* gdb.base/hbreak.c: New file.

--- gdbserver/i386-low.h	2009-06-30 02:09:41.000000000 -0700
+++ gdbserver/i386-low.h	2009-06-30 03:35:55.000000000 -0700
@@ -82,6 +82,16 @@ extern int i386_low_stopped_data_address
 /* Return true if the inferior has some watchpoint that triggered.
    Otherwise return false.  */
 extern int i386_low_stopped_by_watchpoint (struct i386_debug_reg_state *state);
+
+/* Insert a hardware-assisted breakpoint at ADDR.
+   Return 0 on success, -1 on failure.  */
+extern int i386_low_insert_hw_breakpoint (struct i386_debug_reg_state *state,
+					  CORE_ADDR addr);
+
+/* Remove a hardware-assisted breakpoint at ADDR.
+   Return 0 on success, -1 on failure.  */
+extern int i386_low_remove_hw_breakpoint (struct i386_debug_reg_state *state,
+					  CORE_ADDR addr);
 \f
 /* Each target needs to provide several low-level functions
    that will be called to insert watchpoints and hardware breakpoints
--- gdbserver/i386-low.c~	2009-06-30 02:10:04.000000000 -0700
+++ gdbserver/i386-low.c	2009-06-30 01:52:51.000000000 -0700
@@ -575,3 +575,37 @@ i386_low_stopped_by_watchpoint (struct i
   CORE_ADDR addr = 0;
   return i386_low_stopped_data_address (state, &addr);
 }
+\f
+/* Support for h/w breakpoints.  */
+
+/* Insert a hardware-assisted breakpoint at ADDR.
+   Return 0 on success, -1 on failure.  */
+
+int
+i386_low_insert_hw_breakpoint (struct i386_debug_reg_state *state,
+			       CORE_ADDR addr)
+{
+  unsigned len_rw = i386_length_and_rw_bits (1, hw_execute);
+  int retval = i386_insert_aligned_watchpoint (state, addr, len_rw);
+
+  if (debug_hw_points)
+    i386_show_dr (state, "insert_hwbp", addr, 1, hw_execute);
+
+  return retval;
+}
+
+/* Remove a hardware-assisted breakpoint at ADDR.
+   Return 0 on success, -1 on failure.  */
+
+int
+i386_low_remove_hw_breakpoint (struct i386_debug_reg_state *state,
+			       CORE_ADDR addr)
+{
+  unsigned len_rw = i386_length_and_rw_bits (1, hw_execute);
+  int retval = i386_remove_aligned_watchpoint (state, addr, len_rw);
+
+  if (debug_hw_points)
+    i386_show_dr (state, "remove_hwbp", addr, 1, hw_execute);
+
+  return retval;
+}
--- gdbserver/linux-x86-low.c	2009-06-30 02:08:55.000000000 -0700
+++ gdbserver/linux-x86-low.c	2009-06-30 01:50:05.000000000 -0700
@@ -433,6 +433,9 @@ x86_insert_point (char type, CORE_ADDR a
   struct process_info *proc = current_process ();
   switch (type)
     {
+    case '1':
+      return i386_low_insert_hw_breakpoint (&proc->private->arch_private->debug_reg_state,
+					    addr);
     case '2':
     case '3':
     case '4':
@@ -450,6 +453,9 @@ x86_remove_point (char type, CORE_ADDR a
   struct process_info *proc = current_process ();
   switch (type)
     {
+    case '1':
+      return i386_low_remove_hw_breakpoint (&proc->private->arch_private->debug_reg_state,
+					    addr);
     case '2':
     case '3':
     case '4':


Index: testsuite/gdb.base/hbreak.c
===================================================================
RCS file: testsuite/gdb.base/hbreak.c
diff -N testsuite/gdb.base/hbreak.c
--- /dev/null	1 Jan 1970 00:00:00 -0000
+++ testsuite/gdb.base/hbreak.c	30 Jun 2009 10:38:35 -0000
@@ -0,0 +1,28 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+   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/>.  */
+
+void
+break_function (void)
+{
+}
+
+int
+main ()
+{
+  break_function ();
+  return 0;
+}
Index: testsuite/gdb.base/hbreak.exp
===================================================================
RCS file: testsuite/gdb.base/hbreak.exp
diff -N testsuite/gdb.base/hbreak.exp
--- /dev/null	1 Jan 1970 00:00:00 -0000
+++ testsuite/gdb.base/hbreak.exp	30 Jun 2009 10:38:35 -0000
@@ -0,0 +1,69 @@
+#   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/>.
+
+# Test hbreak.
+# Targets need to add "set_board_info gdb,hardware_breakpoints 1"
+# to their board files.
+
+if $tracelevel {
+    strace $tracelevel
+}
+
+# Skip if not supported.
+
+if ![target_info exists gdb,hardware_breakpoints] {
+    unsupported "hbreak"
+    return
+}
+
+set testfile "hbreak"
+set srcfile ${testfile}.c
+set binfile ${objdir}/${subdir}/${testfile}
+
+if  { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } {
+    untested hbreak.exp
+    return
+}
+
+if [get_compiler_info ${binfile}] {
+    return
+}
+
+# Start with a fresh gdb.
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load $binfile
+
+# We can't do hbreaks yet as the "exec" target doesn't support hbreak.
+# Once we get the x86 linux child target instantiated, for example,
+# then we can do hbreak.  So first run to main.
+# ??? Arguably broken, not sure how to fix.
+
+if { ![runto_main] } {
+    fail "Can't run to main"
+    return
+}
+
+gdb_test "hbreak break_function" \
+         "Hardware assisted breakpoint 2 at .*: file .*${srcfile}, line .*" \
+         "breakpoint on break_function"
+
+gdb_test "continue" \
+         ".*Breakpoint 2, break_function ().*" \
+         "run to break_function"
+
+gdb_continue_to_end "hbreak"


             reply	other threads:[~2009-06-30 10:54 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-06-30 10:54 Doug Evans [this message]
2009-07-01  8:34 ` Pierre Muller

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=20090630105351.5FBD7846CA@localhost \
    --to=dje@google.com \
    --cc=gdb-patches@sourceware.org \
    /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