[lldb-dev] expected timeouts and reruns

[Todd Fiala via lldb-dev]

Hi all,

If you happen to use --rerun-all-issues to turn on test rerunning (via
single worker thread) for any failed issue, one thing to be aware of is
that expected timeouts that do time out will not be rerun.  They are not
eligible for rerun (at least as of r255641) since they wouldn't cause the
test run to fail.  (I just hit this and thought I'd note it since it might
not be intuitive).  We only rerun things that would cause a test run to
otherwise fail.

One way to deal with this is to not have these tests marked as expected
timeout, but that'll only be useful if the test would be eligible for
rerun.  Right now rerun eligibility is controlled by either directly
marking a test as flakey or using the --rerun-all-issues command line
argument.

I expect we'll likely discuss and finesse this as we go forward.

-- 
-Todd
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/lldb-dev/attachments/20151215/bd9c73c6/attachment.html>