BIND 10 #542: catching exceptions in run_unittests
BIND 10 Development
do-not-reply at isc.org
Fri May 20 22:33:14 UTC 2011
#542: catching exceptions in run_unittests
-------------------------------------+-------------------------------------
Reporter: | Owner: jinmei
jinmei | Status: reviewing
Type: | Milestone:
enhancement | Sprint-20110531
Priority: minor | Resolution:
Component: build | Sensitive: 0
system | Sub-Project: DNS
Keywords: | Estimated Difficulty: 3.0
Defect Severity: N/A | Total Hours: 0
Feature Depending on Ticket: |
Add Hours to Ticket: 0 |
Internal?: 0 |
-------------------------------------+-------------------------------------
Comment (by jinmei):
Overall it looks good. I've noticed some points that could be further
improved, made these changes in the repository and pushed them. I
believe these are trivial, so please directly check the diff. Of
course, if you don't agree with any of them please feel free to
complain.
There are a couple of more points we might want to discuss:
- maybe a matter of taste, but it seems we can now merge tests and
io_tests. wouldn't it be simpler?
- Also the ordering is still a bit tricky, but with your change I'd
not consider it a "hack". I'd rephrase the comment like this:
{{{
SUBDIRS = . io unittests tests io_tests
# Note the position of unittests: It uses io and tests and io_tests use
# unittest.
}}}
(if we merge tests and io_tests the comment should be adjusted
accordingly)
If you agree these make sense, please change it. If not, I'm okay
with that, too.
And, there's another thing that may be controversial: as mentioned in
the ticket, I'd like to enable it by default for clang++. In fact,
there doesn't seem to be a reason for not enabling it for clang++
because it doesn't print exceptions in a fancy format or keep deeper
stack frames. I'm attaching a proposed patch for that.
One more thing: I'd also suggest introducing a wrapper for
EXPECT_NO_THROW as mentioned in:
https://lists.isc.org/pipermail/bind10-dev/2011-January/001891.html
but the amount of diff would be even bigger, so we should probably
defer it to a separate ticket (I'm completely fine with that).
Finally, you may or may not want to update ChangeLog for this (I'm
okay with either way). If you do, please provide proposed text.
--
Ticket URL: <http://bind10.isc.org/ticket/542#comment:11>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development
More information about the bind10-tickets
mailing list