Path: utzoo!censor!geac!torsqnt!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!mcsun!unido!ira.uka.de!ifistg!ifi!raichle
From: raichle@azu.informatik.uni-stuttgart.de (Bernd Raichle)
Newsgroups: comp.sys.atari.st
Subject: Re: TOS 1.4 pinhead, FATSPEED
Message-ID: <RAICHLE.91Jan23122021@azu.informatik.uni-stuttgart.de>
Date: 23 Jan 91 11:20:21 GMT
References: <14597@milton.u.washington.edu>
	<1991Jan18.212946.24607@opusc.csd.scarolina.edu>
	<1991Jan22.160120.16423@cbnewsl.att.com>
Sender: news@ifistg.uucp
Organization: IfI, Univ. Stuttgart, W Germany
Lines: 17
In-reply-to: hojo@cbnewsl.att.com's message of 22 Jan 91 16:01:20 GMT

In article <1991Jan22.160120.16423@cbnewsl.att.com>, hojo@cbnewsl.att.com (HC Johnson) writes:

] 2.
] pinhead allows you to take advantage of a new feature in TOS 1.4; namely
] whether to ZERO the Bss or not.
]
] Like MS/DOS TOS normally clears ALL unused memory that is initially assigned
] to a process.  On multi-meg processors this is somewhat slow.  (Again TOS 1.4
] uses better code and does clear faster.).  There is a new bit in the process
] header that indicates that TOS need not clear bss, as the process will do its
] own initialization.

I think, the BSS section is (and should be) always cleared!
With the "fastload" bit set, TOS 1.4 doesn't clear the heap (= memory
which could be allocated with malloc() calls).

-bernd