From: dontmailme AT iname DOT com (Steamer)
Newsgroups: comp.os.msdos.djgpp
Subject: Re: Datatype sizes
Date: Tue, 09 May 2000 16:11:24 GMT
Organization: always disorganized
Lines: 11
Message-ID: <39183889.1099585@news.freeserve.net>
References: <Pine DOT SUN DOT 3 DOT 91 DOT 1000509120427 DOT 946M-100000 AT is> <200005091019 DOT MAA17313 AT maggiore DOT iperbole DOT bologna DOT it> <391812bf DOT 18389559 AT news DOT freeserve DOT net> <n0dghs4cses9ehc6iq728lt0dv4pg1glsv AT 4ax DOT com>
NNTP-Posting-Host: modem-77.south-carolina.dialup.pol.co.uk
X-Trace: newsg3.svr.pol.co.uk 957888684 4262 62.137.91.77 (9 May 2000 16:11:24 GMT)
NNTP-Posting-Date: 9 May 2000 16:11:24 GMT
X-Complaints-To: abuse AT theplanet DOT net
X-Newsreader: Forte Free Agent 1.11/32.235
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

Damian Yerrick wrote:

>Anyone know why `short' variables are said to be so much slower than
>`int'/`long' variables, even if they don't cross boundaries?

In 32-bit code there is a special prefix byte needed before any
16-bit instruction.  This can take an extra clock cycle to decode
(although this cycle may disappear in the pipeline on some
processors, depending on the surrounding code).

S.