From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me X-Spam-Level: X-Spam-Status: No, score=-1.3 required=5.0 tests=BAYES_00,INVALID_MSGID autolearn=no autolearn_force=no version=3.4.4 X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 103376,bdd927ef8bb6088d X-Google-Attributes: gid103376,public From: kenner@lab.ultra.nyu.edu (Richard Kenner) Subject: Re: Announcing GNAT version 3.12p for Linux and Sparc Solaris Date: 1999/10/30 Message-ID: #1/1 X-Deja-AN: 542503107 References: <231019991644109306%hopperj@macconnect.com> <3812878f_4@news1.prserv.net> <241019991113025768%hopperj@macconnect.com> X-Trace: typhoon.nyu.edu 941311404 128.122.140.194 (Sat, 30 Oct 1999 15:23:24 EDT) Organization: New York University Ultracomputer Research Lab NNTP-Posting-Date: Sat, 30 Oct 1999 15:23:24 EDT Newsgroups: comp.lang.ada Date: 1999-10-30T00:00:00+00:00 List-Id: In article <241019991113025768%hopperj@macconnect.com> "James E. Hopper" writes: >i am refering to the Machten/Codebuilder port. the problem with ppc >linux is that gcc will not boot strap (c or ada) unless you build it >with either egcs or gcc 2.95 neither of which is supported for gnat. >someone has patches to replace the gcc-281.dif patche that come with >gnat, but i havent been able to find them. I don't believe they exist as it is a considerable amount of work to create them. Moreover, those patches constitute the *easiest* of the three tasks needed to have GNAT and GCC 2.95 interoperate. Right now, I'm working on the task of merging in patches to the GCC2 source tree made after the last EGCS merge back into the GCC development tree. This is a very nontrivial task because of changes made during EGCS development. About 60% of the patches go in unchanged, about half of the remaining ones represent bugs fixed independently by EGCS, and the rest need changes, some considerable, to apply to GCC 2.95. I've been working on this for the last couple of days and estimate I'm half done. That effort would produce the a set of patches that are similar to what you've asked for. But there are two much more difficult tasks that also need to be done: (1) Modify GNAT (specifically Gigi) to reflect changes in GCC infrastructure and the front-end interface made since 2.8.1. There are a set of patches to do some of this (perhaps these are what you recall seeing), but they are far from complete at this point. (2) Debug the resulting compiler by runing ACVC and various other test suites. This will undoubtably uncover not just bugs in the two steps above but also bugs in other GCC changes (including the new optimizers) that only affect Ada code. Having GNAT work with the latest GCC sources is something that a lot of people want to see happen, including Ada Core Technologies, the GCC Steering Committee, and the Ada/Linux group, so it it definitely happen, but it will be a lot of work and take a lot of time. The work I mention above is definitely the first step in that task, though.