From: David Edelsohn <dje@watson.ibm.com>
To: Mark Mitchell <mark@codesourcery.com>
Cc: Michael Matz <matz@suse.de>, Geoff Keating <geoffk@geoffk.org>,
"mrs@apple.com" <mrs@apple.com>, "steby@enea.se" <steby@enea.se>,
"gcc@gcc.gnu.org" <gcc@gcc.gnu.org>
Subject: Re: GCC floating point usage
Date: Wed, 16 Oct 2002 15:27:00 -0000 [thread overview]
Message-ID: <200210162134.RAA24036@makai.watson.ibm.com> (raw)
In-Reply-To: Message from Mark Mitchell <mark@codesourcery.com> of "Wed, 16 Oct 2002 14:05:02 PDT." <364550000.1034802302@warlock.codesourcery.com>
>>>>> Mark Mitchell writes:
Mark> We're still hung up on a more basic problem, though; the RS6000
Mark> maintainers haven't yet made an up-or-down decision on the idea of
Mark> the flag...
I do not fundamentally object to a commandline option, for some
set of semantics which remains ambigous in this discussion, IMHO. I would
prefer if GCC implicitly avoided referencing FPRs in functions that do not
otherwise use floating point computations instead of an explicit option.
I believe there are problems with the implementations of the flag
presented so far, both from a GCC Internals perspective (e.g., reload) and
a code generation perspective (e.g., emitting two successive GPR load or
store instructions which are not scheduled).
I am reserving final judgment until I see a complete definition of
the semantics and an implementation that is more universally accepted as
correct. Neither Geoff nor I want to take on the maintenance burden of a
flawed patch.
David
next prev parent reply other threads:[~2002-10-16 21:35 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-10-14 9:59 Stefan Bylund
2002-10-14 10:16 ` David Edelsohn
2002-10-14 10:25 ` Zack Weinberg
2002-10-14 10:55 ` Dale Johannesen
2002-10-14 22:22 ` Mark Mitchell
2002-10-15 15:35 ` Geoff Keating
2002-10-15 16:10 ` Mark Mitchell
2002-10-15 17:20 ` Geoff Keating
2002-10-15 18:09 ` Mark Mitchell
2002-10-16 7:40 ` Joel Sherrill
2002-10-15 19:04 ` Mike Stump
2002-10-16 12:06 ` Mark Mitchell
2002-10-16 13:35 ` Geoff Keating
2002-10-16 14:29 ` Mark Mitchell
2002-10-16 14:56 ` Michael Matz
2002-10-16 15:03 ` Mark Mitchell
2002-10-16 15:27 ` David Edelsohn [this message]
2002-10-16 15:36 ` Mark Mitchell
2002-10-16 16:35 ` Zack Weinberg
2002-10-16 16:36 ` Mark Mitchell
2002-10-16 16:46 ` David Edelsohn
2002-10-17 8:37 ` Paul Koning
2002-10-16 17:57 ` Mike Stump
2002-10-17 4:12 ` Mike Stump
2002-10-16 13:43 ` Richard Henderson
2002-10-16 14:35 ` Mark Mitchell
2002-10-16 14:56 ` Joel Sherrill
2002-10-16 16:38 ` Richard Henderson
2002-10-16 16:53 ` Zack Weinberg
2002-10-16 17:52 ` Michael Matz
2002-10-16 22:50 ` Richard Henderson
2002-10-21 12:21 ` Jeff Law
2002-10-16 17:29 ` Mike Stump
2002-10-17 2:19 ` Richard Henderson
2002-10-15 17:19 ` Mike Stump
2002-10-15 18:41 ` Zack Weinberg
2002-10-16 1:48 ` Fergus Henderson
2002-10-14 10:37 ` Stefan Bylund
2002-10-14 11:28 ` Mike Stump
2002-10-14 12:39 ` Joel Sherrill
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=200210162134.RAA24036@makai.watson.ibm.com \
--to=dje@watson.ibm.com \
--cc=gcc@gcc.gnu.org \
--cc=geoffk@geoffk.org \
--cc=mark@codesourcery.com \
--cc=matz@suse.de \
--cc=mrs@apple.com \
--cc=steby@enea.se \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).