Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/26/83; site ihuxm.UUCP
Path: utzoo!linus!decvax!microsoft!uw-beaver!cornell!vax135!ariel!houti!hogpc!houxm!hocda!spanky!burl!we13!ihnp4!ihuxm!rck
From: rck@ihuxm.UUCP
Newsgroups: net.lang.c
Subject: Re: Whitesmith and static variable init.
Message-ID: <364@ihuxm.UUCP>
Date: Fri, 15-Jul-83 12:30:27 EDT
Article-I.D.: ihuxm.364
Posted: Fri Jul 15 12:30:27 1983
Date-Received: Tue, 19-Jul-83 12:55:43 EDT
References: <413@ut-ngp.UUCP>, <140@ihu1f.UUCP>, <792@umcp-cs.UUCP>
Organization: BTL Naperville, Il.
Lines: 15

I've had some experience with Whitesmith's C compiler and support tools.
Unfortunately, their variable initialization does not follow the accepted
standard, but they don't have much of a choice.  The Whitesmith package
is designed for use on small "operating systems" like CP/M.  CP/M does
not support any kind of runtime loader that zeroes a program's bss
section, so apparently Whitesmith's has taken the approach of forcing
all statics to be initialized in order to force them into the data
section of the program.  So in this way, the initialized variables, be
they scalar or array, are part of the program image and are copied in
when the program is run.  I don't know why Whitesmith's just didn't put
ALL statics into the data section, even if they weren't explicitly
initialized.  Maybe someone else could respond.
						Ron Kukuk
						Bell Labs
						Naperville