ia64 libjava java-signal.h build failure

[Mark Mitchell]

>>>>> "Alexandre" == Alexandre Oliva <aoliva@redhat.com> writes:

    Alexandre> They can always disable the java language.

Again, that misses the point: these people shouldn't have to suffer
that, especially not on the release branch at this stage.  It's our
responsibility to test this first.

    Alexandre> by default, we just won't learn about it, and the
    Alexandre> release will be broken for those who try to enable
    Alexandre> java.

Not if you do as I asked.  All I'm asking is that you test the feature
you want to enable, across a range of platforms, before you enable it.

Under your proposal, other people have to do that testing instead, and
if something goes wrong, they suffer.

I feel that I've offerred a sensible way to accomplish your goal.  All
it requires is a little effort on your part, which seems fair since
you are the one who wants the feature enabled.  Prepare the patch, run
the tests, report the results -- and then we can check it in.

I'm happy to raise this with the SC if you'd like.  Thankfully for
everyone, I am *never* the final authority.

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