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,e29c511c2b08561c X-Google-Attributes: gid103376,public From: dewar@cs.nyu.edu (Robert Dewar) Subject: Re: Is the "Ada mandate" being reconsidered? Date: 1996/06/30 Message-ID: #1/1 X-Deja-AN: 163656160 references: <9606212019.AA11075@eight-ball> <4qqc4s$flv@felix.seas.gwu.edu> organization: Courant Institute of Mathematical Sciences newsgroups: comp.lang.ada Date: 1996-06-30T00:00:00+00:00 List-Id: Ronald Cole said "And those few problems that you do have (-O2 generates bad code on HPPA) don't look like they're going to be fixed any time soon..." On the contrary, gcc problems that affect supported customers have been fixed pretty rapidly, we treat them like any other problems. However, the particular situation you refer to (-O2 on the HPPA) is not in this category since, as I assume you know, we are not yet offering support for the HPPA version of GNAT. We are planning on offering support for the HPPA version of GNAT in the third quarter, and this is one of the problems that needs to be resolved before we offer such support. Right now, this is certainly not a high priority problem for us, since our priorities are dictated by our supported customers. That being said, we do have a lot of successful users of the HPPA version right now, and some large projects have been successfully completed using the HPPA version. Most of these projects (at least those we know of), have not been significantly affected by the bug you refer to, although it does (as noted in the readme file) affect the bootstrap. We still have to bootstrap the HPPA version at -O1. Note that there are some other critical issues with the HPPA port before we can offer full support. In particular, we have to get tasking fully operational, which is non-trivial, since DCE threads are quite broken from Ada's point of view. We have figured out how to implement full Ada tasking under DCE threads, but it will take a bit of fiddling around with the runtime and he compiler -- which we are working on right now, since there are other targets that will need to use the DCE threads interface.