X-pop3-spooler:	POP3MAIL 2.1.0 b 3 961213 -bs-
Delivered-To:	pcg AT goof DOT com
Message-ID:	<19980312234128.63392@cerebro.laendle>
Date:	Thu, 12 Mar 1998 23:41:28 +0100
From:	Marc Lehmann <pcg AT goof DOT com>
To:	Thomas Koehler <T DOT Koehler AT pfh DOT research DOT philips DOT com>
Cc:	beastium <beastium-list AT Desk DOT nl>
Subject:	Re: paranoia & extra precision [was -fno-float-store in pgcc]
References:	<199803111756 DOT AA192209001 AT typhoon DOT rose DOT hp DOT com> <19980312074656 DOT 14562 AT cerebro DOT laendle> <35081A31 DOT 16F7D911 AT pfh DOT research DOT philips DOT com>
Mime-Version:	1.0
X-Mailer:	Mutt 0.88
In-Reply-To:	<35081A31.16F7D911@pfh.research.philips.com>; from Thomas Koehler on Thu, Mar 12, 1998 at 06:24:01PM +0100
X-Operating-System:	Linux version 2.1.85 (root AT cerebro) (gcc version pgcc-2.91.06 980129 (gcc-2.8.0 release))
Status:	RO
Lines:	23

On Thu, Mar 12, 1998 at 06:24:01PM +0100, Thomas Koehler wrote:
> > As all of you correctly guessed, this is helplessly slow, but the only way
> > to force correct behaviour. 
> 
> There seems to be annother way, but involes the unducumented functions
> __getfpucw and __setfpucw.

undocumented?? anyway, these still won't make your program ieee compliant.

- not all operations are affected
- the compiler cannot know when to set it, i.e. the same fundamental problem
  than with mmx, the compiler had to set the cw on almost every procedure
  entry etc...

x86 fpu = hopelessly broken. by design.

      -----==-                                              |
      ----==-- _                                            |
      ---==---(_)__  __ ____  __       Marc Lehmann       +--
      --==---/ / _ \/ // /\ \/ /       pcg AT goof DOT com       |e|
      -=====/_/_//_/\_,_/ /_/\_\                          --+
    The choice of a GNU generation                        |
                                                          |

- Raw text -