testsuite fixes for bleeding edge gcc

[Zack Weinberg]

Michael Chastain <mec.gnu@mindspring.com> writes:

>> whether my GCC changes worked.
>
> Ah.  That is a good reason for me to keep up maintenance on the
> 6.2 branch, if it happens often enough.  Just because we don't
> release from the branch doesn't mean it's useless.

I don't myself have plans to do this on a regular basis, but it would
be nice if it were something that GCC developers could do easily,
because (as you know) GCC's own testsuite doesn't catch most bugs in
debugging-info output.

This and the use of KFAIL (forcing me to upgrade dejagnu) were the
only major problems I had.  

zw