Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm
Sender: cygwin-owner AT sourceware DOT cygnus DOT com
Delivered-To: mailing list cygwin AT sourceware DOT cygnus DOT com
Message-Id: <199905030225.VAA14434@modi.xraylith.wisc.edu>
X-Authentication-Warning: modi.xraylith.wisc.edu: localhost.xraylith.wisc.edu [127.0.0.1] didn't use HELO protocol
To: Chris Faylor <cgf AT cygnus DOT com>
cc: cygwin AT sourceware DOT cygnus DOT com
Subject: Re: (patch) newlib rint fix 
In-reply-to: Your message of "Sun, 02 May 1999 19:45:43 EDT."
             <19990502194542 DOT C950 AT cygnus DOT com> 
Date: Sun, 02 May 1999 21:25:52 -0500
From: Mumit Khan <khan AT xraylith DOT wisc DOT EDU>

Chris Faylor <cgf AT cygnus DOT com> writes:
> 
> Ok.  I wasn't necessarily advocating the use of -ffloat-store, I was
> just relaying the one response to the "volatile" fix that you mentioned.

I understand. I just wanted to point out the drawbacks of float-store.
I've had to add tons of "weird" tweaks to my own numerical code to avoid 
these problems[1]. It's useful to note that the same problems occur with
gcc/egcs on other x86 OS's as well, even if in different contexts. 

> It's entirely likely that we might (soon) have an assembly fix for this,
> similar to Jacob Navia's.  Hopefully, the newlib maintainer will have a
> strong opinion on how this should be fixed and will fix it soon once and
> for all.

Great. Thanks for the heads up.

Regards,
Mumit

[1] These hacks aren't necessary using a few non-free compilers that I 
use, where the rounding modes are dealt with internally somehow. There 
are still problems however, notably when using 128bit doubles and such.


--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe AT sourceware DOT cygnus DOT com