>> I actually meant a negative test case. > > As I imagine you saw, there's already one there that uses speculative > tracing and verifies that enablings are not reaped: > > usr/src/cmd/dtrace/test/tst/common/usdt/tst.noreap.ksh > > Having a test that uses ring buffering is certainly possible, it's > just a more complicated test to write. (I'm currently cheating a bit > by using one enabling to kick off another.) But I'm happy to add it > if you'd like to see it...
Yes. I saw that test case. It was just a suggestion to exercise the other path in your code, but it's up to you. >> Would that be worth a one- or two-line comment? > > There's already the comment that it's padded out to avoid false > sharing; does it merit more than that? I guess that's sufficient, but I needed to remind myself that these were per-CPU data structures. Your call. Adam -- Adam Leventhal, Delphix http://dtrace.org/blogs/ahl 275 Middlefield Road, Suite 50 Menlo Park, CA 94025 http://www.delphix.com _______________________________________________ dtrace-discuss mailing list dtrace-discuss@opensolaris.org
