This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: patch for -Wno-long-long and early GNAT compilers
- From: dewar at gnat dot com (Robert Dewar)
- To: dewar at gnat dot com, mrs at windriver dot com, zack at codesourcery dot com
- Cc: gcc at gcc dot gnu dot org, kenner at vlsi1 dot ultra dot nyu dot edu
- Date: Mon, 18 Mar 2002 21:02:26 -0500 (EST)
- Subject: Re: patch for -Wno-long-long and early GNAT compilers
> From my point of view, what is considered to be a justifiable use of
> ACT's company resources is irrelevant.
Well it may be so from your point of view, but the facts of the matter are
as follows.
ACT can provide a limited amount of resources for helping to maintain and
update the GCC sources in the FSF tree. No amount of yelling or anything
else will increase this resource. The important thing is to use this limited
resource most effectively.
With my FSF maintainer's hat on, I think the priorities from FSF's point
of view are:
1. Have the latest best version of the sources synchronized at gnu.org
2. Make sure this latest best version is reliable testing against all the
various test suites.
3. Fixing bugs
I must say that Zack's concern (bootstrapping with old versions of the
compiler or odd non-standard mixed versions of GNAT) is not even on my
radar screen of priorities.
What is encouraging to me is that we are making really good progress on
points 1 and 2. The sources are reasonably well sychronized now, and we
are gradually solving the problem of automating the process of keeping
them better synchronized (e.g. we just solved today the issue of version
numbers that we need for our configuration control at ACT, but which are
not wanted or needed in the FSF tree -- note that the point here is that
we want to avoid this issue of version numbers being a drain on our
limited resources. We want to be doing useful stuff, not fiddling with
version numbers -- and we finally figured out a smooth way of dealing
with this so that all goals can be met without significant fuss.
With regard to point 2, we are now completely clean on Solaris, and very
close to clean (one code generaiton problem left) on GNU/Linux. We are
now moving on to the next step, which is getting things working well on
AIX, HPUX, and Tru64.
I really think we are making real progress here, and we are close to having
a realiable GNAT 5 build. We still have to deal with all the tools, so we
have more to do, but we are definitely making progress.