summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
-rw-r--r--lldb/test/lang/cpp/class_types/TestClassTypes.py11
1 files changed, 9 insertions, 2 deletions
diff --git a/lldb/test/lang/cpp/class_types/TestClassTypes.py b/lldb/test/lang/cpp/class_types/TestClassTypes.py
index 2012b889144..4be117a0d0a 100644
--- a/lldb/test/lang/cpp/class_types/TestClassTypes.py
+++ b/lldb/test/lang/cpp/class_types/TestClassTypes.py
@@ -159,8 +159,12 @@ class ClassTypesTestCase(TestBase):
# Is this a case of clang (116.1) generating bad debug info?
#
# Break on the ctor function of class C.
- self.expect("breakpoint set -M C", BREAKPOINT_CREATED,
- startstr = "Breakpoint created: 1: name = 'C'")
+ #self.expect("breakpoint set -M C", BREAKPOINT_CREATED,
+ # startstr = "Breakpoint created: 1: name = 'C'")
+
+ # Make the test case more robust by using line number to break, instead.
+ self.expect("breakpoint set -l %d" % self.line, BREAKPOINT_CREATED,
+ startstr = "Breakpoint created")
self.runCmd("run", RUN_SUCCEEDED)
@@ -174,6 +178,7 @@ class ClassTypesTestCase(TestBase):
substrs = [' resolved, hit count = 1'])
# Continue on inside the ctor() body...
+ self.runCmd("register read pc")
self.runCmd("thread step-over")
# Verify that 'frame variable this' gets the data type correct.
@@ -181,6 +186,8 @@ class ClassTypesTestCase(TestBase):
substrs = ['C *'])
# Verify that frame variable -T this->m_c_int behaves correctly.
+ self.runCmd("register read pc")
+ self.runCmd("expr m_c_int")
self.expect("frame variable -T this->m_c_int", VARIABLES_DISPLAYED_CORRECTLY,
startstr = '(int) this->m_c_int = 66')
OpenPOWER on IntegriCloud