[08/10] watchpoint.exp: Expose kfail gdb/38 (PR7143) again.

[Pedro Alves]

PR7143 is about a watchpoint not triggering when first set.  There's a
KFAIL in the test file to exercise the bug, but it stopped triggering by
accident at some point, because the test's procedure runs with hw watchpoints
force disabled.  This fixes it, re-exposing the bug.  It also makes
several other test run with hw watchpoints enabled that were never
tested that way before, most certainly not by intention.

2012-01-26  Pedro Alves  <palves@redhat.com>

	* gdb.base/watchpoint.exp (top level): Re-enable hardware
	watchpoints sooner.
---

 gdb/testsuite/gdb.base/watchpoint.exp |   10 +++++-----
 1 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/gdb/testsuite/gdb.base/watchpoint.exp b/gdb/testsuite/gdb.base/watchpoint.exp
index 8c5e1ee..697f46e 100644
--- a/gdb/testsuite/gdb.base/watchpoint.exp
+++ b/gdb/testsuite/gdb.base/watchpoint.exp
@@ -876,6 +876,11 @@ if [initialize] then {
 	"Expression cannot be implemented with read/access watchpoint..*" \
 	"rwatch disallowed when can-set-hw-watchpoints cleared"

+    # Re-enable hardware watchpoints if necessary.
+    if ![target_info exists gdb,no_hardware_watchpoints] {
+        gdb_test_no_output "set can-use-hw-watchpoints 1" ""
+    }
+
     test_watchpoint_and_breakpoint

     test_watchpoint_in_big_blob
@@ -886,11 +891,6 @@ if [initialize] then {

     test_watch_location

-    # Re-enable hardware watchpoints if necessary.
-    if ![target_info exists gdb,no_hardware_watchpoints] {
-        gdb_test_no_output "set can-use-hw-watchpoints 1" ""
-    }
-
     test_wide_location_1
     test_wide_location_2
 }