Re: F77 makefile again.

[craig at jcb-sc dot com]

>Why do I have an impression that people who modify g77 makefiles in
>egcs either don't care about parallel make or don't want to spend
>time to write good makefiles.

Well, part of the problem is that we don't test parallel make (I doubt
my poor P2/233 could make much use of it), and we basically never get
any proper bug reports showing failures during parallel make.  (The
occasional bug report we do get just doesn't tell us enough to help
us understand the problem.)  Also, we know we're going to be rewriting
the whole configury mess anyway, and, as Jeff Law seems to also feel,
our learning the ins and outs of `make' is hardly productive when our
job is to work on the guts of the compiler(s).

In *this* case, though, it's purely my fault.  While working on my
recent patch, the thought did fly through my head that, perhaps,
there was a dependency I should list to ensure that $libsubdir had
been made, but I didn't nailed that thought down to the inside of
my skull long enough to ensure it made it into my patch after
a very quick bit of research (which I've since done).

A fix will be forthcoming.  Thanks for pointing it out!  Though, I
assume, it won't really fix parallel makes involving g77, given all
the other problems you have claimed exist in the past, it's at least
nice to avoid introducing new problems.

>There is no dependency for f77.install-common. With parallel make, it
>may fail since $(libsubdir) may not exit when f77.install-common is
>reached.

Yup, need dependencies on `installdirs'.

        tq vm, (burley)