Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!ncar!husc6!endor!siegel
From: siegel@endor.harvard.edu (Rich Siegel)
Newsgroups: comp.sys.mac.programmer
Subject: Re: Bug in LSP V2.01?
Message-ID: <2925@husc6.harvard.edu>
Date: 21 Oct 89 03:59:07 GMT
References: <3935@ethz-inf.UUCP>
Sender: news@husc6.harvard.edu
Reply-To: siegel@endor.UUCP (Rich Siegel)
Organization: Symantec Language Products Group
Lines: 41

In article <3935@ethz-inf.UUCP> schorn@inf.ethz.ch (Peter Schorn) writes:
>
> procedure obj.init;
>  procedure h(var y:integerP); begin new(y); y^:=19; end;
> begin h(x); end;
>
> var o: obj;
>begin new(o); o.init; writeLn(o.x^); end.
>
>Is this a compiler bug or am I doing something wrong?

I'd venture to guess that you're doing something wrong. Since objects
are allocated as handles (relocatable blocks), passing a field of an
object as a VAR parameter to a routine that may move memory (which
procedure H does, because New() calls NewPtr(), which moves memory)
is a distinctly unsafe thing to do. Before calling procedure H,
method O.Init should HLock(Handle(SELF)), and HUnlock(Handle(SELF))
after H returns.

(That's a simplification; actually, you'll want to HGetState(),
move SELF high, lock it, and then HSetState when you're done.)

You may wish to define and implement a TObject.Lock() and TObject.Unlock
for convenience.

>Peter                                     schorn@inf.ethz.ch  or
>                                          schorn@cs.unc.edu

		---Rich

~~~~~~~~~~~~~~~
 Rich Siegel
 Staff Software Developer
 Symantec Corporation, Language Products Group
 Internet: siegel@endor.harvard.edu
 UUCP: ..harvard!endor!siegel

"There is no personal problem which cannot be solved by sufficient
application of high explosives."

~~~~~~~~~~~~~~~