[Lldb-commits] [PATCH] Enabled a bunch of tests on Linux

Siva Chandra sivachandra at google.com
Tue Mar 31 04:17:05 PDT 2015

Please do not view my comments as blockers. As indicated in my first post, they are only notes for the trail.

Comment at: test/functionalities/dynamic_value_child_count/TestDynamicValueChildCount.py:24
@@ -23,3 +23,3 @@
-    @expectedFailureLinux # FIXME: This needs to be root-caused.  It looks like the DWARF info is anticipating the derived class assignment.
+    @expectedFailureLinux("llvm.org/pr23039")
     @expectedFailureFreeBSD("llvm.org/pr19311") # continue at a breakpoint does not work
vharron wrote:
> sivachandra wrote:
> > This is failing because of a completely different issue. If you replace the "continue" commands with "next", this test passes. I was digging into this exact test when I saw your patch.
> The build server seems to indicate that this is working on Darwin.  The fact that it doesn't work on Linux concerns me.
What I am trying to hint here is that the test fails, but not because what is being tested is failing. This test case surfaces another, IMO fairly serious, issue.

Comment at: test/python_api/thread/TestThreadAPI.py:78
@@ -77,3 +77,3 @@
     @expectedFailureFreeBSD # llvm.org/pr20476
-    @expectedFailureLinux # llvm.org/pr14416
+    @expectedFailureLinux # need to 'thread step-out' twice to get out of malloc
vharron wrote:
> sivachandra wrote:
> > May be a PR for this as well?
> I will if I have more information than I can fit in the comment.
Your comment seems to be a workaround, which is good enough to be in a PR.



More information about the lldb-commits mailing list