sjlj-exceptions ICE. simplified test case

[Mark Mitchell]

  >From the gcj (Java) point of view, I most strongly object to this
  suggestion.  Yes, we prefer to use range table based exceptions, but
  sjlj exceptions are a useful way to begin porting Java to a new
  target, and to remove sjlj support before we've ported range tables to
  all our targets will cause us great pain.

I understand what you're saying.

I have no strong opinion in this case.

However, we should all be aware that every feature, even ones that
already exist, have a substantial maintenance cost.  Perhaps, from the
point of view gcj development, sjlj exceptions "just work".  But,
Richard just spent some time tracking down bugs that only occurred in
that context, and lots of other people have had to debug similar
problems in the past.  Some of those resources could probably have
been devoted to implementing the DWARF2 unwind support for a few more
targets.

Every time we can reasonably remove a feature, command-line switch,
etc., we should.  That doesn't by any means imply that this is such a
feature, or that we should do so if lots of people depend on the
feature, etc.

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