From: "Michael Beck" <beck AT dresearch DOT de>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: The "long long int" type: is it fast or slow?
Date: Sat, 3 Oct 1998 22:59:35 +0200
Organization: DResearch - Digital Media Systems
Lines: 20
Message-ID: <6v7k8s$ebk$1@latinum.dresearch.de>
References: <PGyJdEAR2QF2EwIl AT lwithers DOT demon DOT co DOT uk>
NNTP-Posting-Host: 195.99.74.37
X-Trace: latinum.dresearch.de 907498588 14708 192.168.0.93 (4 Oct 1998 10:56:28 GMT)
X-Complaints-To: postmaster AT DResearch DOT DE
NNTP-Posting-Date: 4 Oct 1998 10:56:28 GMT
X-Newsreader: Microsoft Outlook Express 4.72.3110.5
X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3
Cache-Post-Path: news-fa!unknown AT 194 DOT 75 DOT 26 DOT 11
X-Cache: nntpcache 2.3.2.1 (see http://www.nntpcache.org/)
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp


Laurence Withers wrote in message ...
>Hi,
>
>Is the GNU C++ "long long int" type (64 bits) fast on the Intel platform
>(using DOS or Windows)? I am mainly interested in bitwise operations (&
>| ! << >> etc) and comparisons (== <= >= !=), rather than arithmetic.
>
>By fast, I mean is it any slower (or even significantly slower) than the
>equivalent operations with 32 bit integers? I couldn't find anything in
>the docs or the FAQ; if I missed something, please point it out.
>
Sure it is slower, because x86 processors didn't support 64 operations. So
in most cases
you need at least 2 CPU operations instead of one to fullfill the operation.

--
Michael Beck   beck AT dresearch DOT de