Re: Floating point differences

Article View: gnu.gcc.help

Article #51716

#51716

From: Hugh Fader
Date: Fri, 16 Apr 1993 16:28

14 lines
1134 bytes

I have just discovered that the epsilon results given in my previous
posting for Microsoft C were with optimization turned on. With
optimization off the results are:

Epsilon machine (avec variable intermediaire) : 1.19209e-007
Epsilon machine (sans variable intermediaire) : 1.0842e-019

So without optimization MSC behaves as cc does.

--
Hugh Fader
Ford Research Laboratory
hugh@slee01.srl.ford.com
(313) 84-58198

Technical Details

Message-ID: <1qmmrtINNb7p@fmsrl7.srl.ford.com>

Path:

arc1-ca.pugleaf.net!gmd.de!newsserver.jvnc.net!news.cac.psu.edu!wintermute.phys.psu.edu!ra!darwin.sura.net!jhunix.hcf.jhu.edu!fmsrl7!hugh

References: <1qmla6INNb25@fmsrl7.srl.ford.com>

Show Full Headers

Complete Article Headers

X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 11318c,e3259ca760e1e814
X-Google-Attributes: gid11318c,public
X-Google-ArrivalTime: 1993-04-16 10:05:40 PST
Path: arc1-ca.pugleaf.net!gmd.de!newsserver.jvnc.net!news.cac.psu.edu!wintermute.phys.psu.edu!ra!darwin.sura.net!jhunix.hcf.jhu.edu!fmsrl7!hugh
From: hugh@slee01.srl.ford.com (Hugh Fader)
Newsgroups: gnu.gcc.help
Subject: Re: Floating point differences
Date: 16 Apr 1993 16:28:45 GMT
Organization: Ford Motor Company Research Laboratory
Lines: 14
Message-ID: <1qmmrtINNb7p@fmsrl7.srl.ford.com>
References: <1qmla6INNb25@fmsrl7.srl.ford.com>
Reply-To: hugh@slee01.srl.ford.com
NNTP-Posting-Host: slee06.srl.ford.com
X-Newsreader: Tin 1.1 PL5

Reply to Parent Message