Re: [3.0 critical] Make fixproto deal with assert.h(bootstrap fails)

[Mark Mitchell <mark at codesourcery dot com>]

>>>>> "Zack" == Zack Weinberg <zackw@Stanford.EDU> writes:

    Zack> I tend to agree with Russ that we can't be expected to
    Zack> support this sort of thing, and that "move it to gcc-lib"
    Zack> may be an adequate answer.

Well, it *is* our own fault, after all.  So, we shouldn't blame the
hapless user.

At the very least we need to detect the situation and tell the user
what to do when they do the configure.  And make sure that it's
something they can do even if they're not root.  I don't know what
this would be, but probably we could come up with something.

I think that would work, but it might be hard.  We migh prefer to
restore __eprintf.

I suppose a final alternative would be to use/install a new assert.h
that did:

  #define __eprintf(...) abort ()
  #include_next <assert.h>

Then, if the installed assert.h is bogus, it would still work.  If the
installed assert.h was OK, our version wouldn't change it.

--
Mark Mitchell                   mark@codesourcery.com
CodeSourcery, LLC               http://www.codesourcery.com