home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.atari.st.tech
- Path: sparky!uunet!cs.utexas.edu!torn!newshost.uwo.ca!uwovax.uwo.ca!7103_2622
- From: 7103_2622@uwovax.uwo.ca (Eric Smith)
- Subject: Re: Can a TSR Malloc()?
- Organization: University of Western Ont, London
- Date: Wed, 12 Aug 1992 13:52:16 GMT
- Message-ID: <1992Aug12.095216.1@uwovax.uwo.ca>
- References: <sourada.713595993@vincent1.iastate.edu>
- Sender: news@julian.uwo.ca (USENET News System)
- Nntp-Posting-Host: hydra.uwo.ca
- Lines: 19
-
- In article <sourada.713595993@vincent1.iastate.edu>, sourada@iastate.edu (Steven D Ourada) writes:
- > Is it possible (and good practice) for a TSR program to Malloc()
- > (the GEMDOS way) after Ptermres()?
-
- No, and no.
-
- Or, more accurately, it's possible, but not likely to produce the desired
- effect. As you point out, GEMDOS will think that it's the invoking program
- that did the Malloc, and will free the memory when the invoker exits.
-
- There is no reliable way to make GEMDOS change the memory ownership of
- Malloc'd memory. The tricks people use to do this *will* break in
- future versions of TOS (in particular, MultiTOS) and neither will they
- work under MiNT.
- --
- Eric R. Smith email:
- Dept. of Mathematics eric.smith@uwo.ca
- University of Western Ontario
-
-