[RFC] Fix verification of changed values for big values.

[RFC] Fix verification of changed values for big values.

[Thiago Jung Bauermann]

Hi,

This is the fix that Sérgio mentioned on the thread about our progress
on the BookE debug registers support.

Right now, GDB calls value_equal when comparing the old and new values
of a watchpoint. IMO this is not correct, since that function will call
coerce_array and effectively just compare the addresses of arrays being
watched.

This patch introduces a new value comparison function which works in the
mentioned case, and a testcase which fails without the patch and passes
with it. Ok to commit?
-- 
[]'s
Thiago Jung Bauermann
IBM Linux Technology Center


gdb/
	* breakpoint.c (value_equal_watchpoint): New function.
	(watchpoint_check): Use value_equal_watchpoint instead
	of value_equal.

gdb/testsuite/
	* gdb.base/watchpoint.exp (test_watchpoint_in_big_blob): New function.
	(top level): Call test_watchpoint_in_big_blob.
	* gdb.base/watchpoint.c (buf): Change size to value too big for hardware
	watchpoints.
	(func3): Write to buf.

Index: gdb.git/gdb/breakpoint.c
===================================================================
--- gdb.git.orig/gdb/breakpoint.c	2009-05-31 17:05:44.000000000 -0300
+++ gdb.git/gdb/breakpoint.c	2009-05-31 17:19:29.000000000 -0300
@@ -2717,7 +2717,23 @@ watchpoints_triggered (struct target_wai
 #define BP_TEMPFLAG 1
 #define BP_HARDWAREFLAG 2
 
-/* Check watchpoint condition.  */
+/* Check watchpoint condition.  We can't use value_equal because it coerces
+   an array to a pointer, thus comparing just the address of the array instead
+   of its contents.  This is not what we want.  */
+
+static int
+value_equal_watchpoint (struct value *arg1, struct value *arg2)
+{
+  struct type *type1, *type2;
+
+  type1 = check_typedef (value_type (arg1));
+  type2 = check_typedef (value_type (arg2));
+
+  return TYPE_CODE (type1) == TYPE_CODE (type2)
+    && TYPE_LENGTH (type1) == TYPE_LENGTH (type2)
+    && memcmp (value_contents (arg1), value_contents (arg2),
+	       TYPE_LENGTH (type1)) == 0;
+}
 
 static int
 watchpoint_check (void *p)
@@ -2785,7 +2801,7 @@ watchpoint_check (void *p)
 
       fetch_watchpoint_value (b->exp, &new_val, NULL, NULL);
       if ((b->val != NULL) != (new_val != NULL)
-	  || (b->val != NULL && !value_equal (b->val, new_val)))
+	  || (b->val != NULL && !value_equal_watchpoint (b->val, new_val)))
 	{
 	  if (new_val != NULL)
 	    {
Index: gdb.git/gdb/testsuite/gdb.base/watchpoint.c
===================================================================
--- gdb.git.orig/gdb/testsuite/gdb.base/watchpoint.c	2009-05-31 17:04:24.000000000 -0300
+++ gdb.git/gdb/testsuite/gdb.base/watchpoint.c	2009-05-31 17:16:35.000000000 -0300
@@ -30,7 +30,7 @@ int ival2 = -1;
 int ival3 = -1;
 int ival4 = -1;
 int ival5 = -1;
-char buf[10];
+char buf[30];
 struct foo
 {
   int val;
@@ -95,6 +95,7 @@ func3 ()
   x = 1;				/* second x assignment */
   y = 1;
   y = 2;
+  buf[26] = 3;
 }
 
 int
Index: gdb.git/gdb/testsuite/gdb.base/watchpoint.exp
===================================================================
--- gdb.git.orig/gdb/testsuite/gdb.base/watchpoint.exp	2009-05-31 17:04:24.000000000 -0300
+++ gdb.git/gdb/testsuite/gdb.base/watchpoint.exp	2009-05-31 17:17:04.000000000 -0300
@@ -678,6 +678,24 @@ proc test_inaccessible_watchpoint {} {
     }
 }
     
+proc test_watchpoint_in_big_blob {} {
+    global gdb_prompt
+
+    gdb_test "watch buf" ".*atchpoint \[0-9\]+: buf"
+    send_gdb "cont\n"
+    gdb_expect {
+	-re "Continuing.*\[Ww\]atchpoint.*buf.*Old value = .*$gdb_prompt $" {
+	    pass "watchpoint on buf hit"
+	}
+	-re "Continuing.*$gdb_prompt $" {
+	    fail "watchpoint on buf hit"
+	}
+	-re ".*$gdb_prompt $" { fail "watchpoint on buf hit" ; return }
+	timeout { fail "watchpoint on buf hit (timeout)" ; return }
+	eof { fail "watchpoint on buf hit (eof)" ; return }
+    }
+}
+
 # Start with a fresh gdb.
 
 gdb_exit
@@ -842,6 +860,8 @@ if [initialize] then {
     }
 
     test_watchpoint_and_breakpoint
+
+    test_watchpoint_in_big_blob
 }
 
 # Restore old timeout

Re: [RFC] Fix verification of changed values for big values.

[Tom Tromey]

>>>>> "Thiago" == Thiago Jung Bauermann <bauerman@br.ibm.com> writes:

Thiago> Right now, GDB calls value_equal when comparing the old and new values
Thiago> of a watchpoint. IMO this is not correct, since that function will call
Thiago> coerce_array and effectively just compare the addresses of arrays being
Thiago> watched.

Thiago> This patch introduces a new value comparison function which works in the
Thiago> mentioned case, and a testcase which fails without the patch and passes
Thiago> with it. Ok to commit?

This makes sense to me.

Thiago> +  return TYPE_CODE (type1) == TYPE_CODE (type2)
Thiago> +    && TYPE_LENGTH (type1) == TYPE_LENGTH (type2)
Thiago> +    && memcmp (value_contents (arg1), value_contents (arg2),
Thiago> +	       TYPE_LENGTH (type1)) == 0;

A style nit: multi-line expressions should be wrapped in parens, and
the continuation lines here should be indented more deeply.

Also, I think value_equal_watchpoint should be in one of the
value-related files.  My reasoning is that I've occasionally run
across code duplication because some function like this was defined in
an obscure place.

Tom