mec.gnu@mindspring.com (Michael Elizabeth Chastain) writes:
> Is there a PR for this bug?  If not, can you file one?  Then change the
> FAIL's to KFAIL's with the PR number on it.  That way I can tell the
> "Invalid control type" from truly unexpected bugs.

I posted the fix when I posted the test.  To make sure things are
tracked properly until the review is done, I've filed gdb/1489, and
committed the first attached patch to make it a kfail.