This is the mail archive of the
mailing list for the GCC project.
Itanium (was gcj 3.0 post-mortem)
- To: "'tromey at redhat dot com'" <tromey at redhat dot com>, Java Discuss List <java at gcc dot gnu dot org>
- Subject: Itanium (was gcj 3.0 post-mortem)
- From: "Boehm, Hans" <hans_boehm at hp dot com>
- Date: Thu, 5 Jul 2001 11:20:37 -0700
- Cc: GCC Hackers <gcc at gcc dot gnu dot org>
> From: Tom Tromey [mailto:firstname.lastname@example.org]
> I believe in the end the IA-64 Linux and Solaris
> ports did not work. This happened because the new EH code required
> more work than we had anticipated, and our lone Java EH expert, Andrew
> Haley, was scheduled to work on other projects during the window we
> needed his expertise.
To correct some previous misstatements on my part:
I would rate gcj3.0 on Linux/IA64 as only semi-broken. It seems to be
usable for some small things. Things that don't work for me:
1) Catching NullPointerExceptions. Andrew pointed out that some other
exception-related things do work with a sufficiently recent linker, which I
didn't have initially.
2) -g1, or any set of options that defaults to -g1, e.g. no compile options.
3) I still have some problems with compiling multiple .class files, though I
doubt those are IA64-related.
I think we have agreement that the first two are really broken in 3.0.
On the positive side, I can build the command-line version of my
constructive reals calculator from source on Itanium with gcj 3.0, and it
seems to run correctly.