Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!ccu.umanitoba.ca!herald.usask.ca!alberta!oha!tony
From: tony@oha.UUCP (Tony Olekshy)
Newsgroups: comp.lang.lisp.x
Subject: Re: signals in XLISP.
Message-ID: <474@oha.UUCP>
Date: 25 Feb 91 18:24:36 GMT
References: <1991Feb13.201943.20045@cs.mcgill.ca> <1170014@hpfcbig.SDE.HP.COM> <1991Feb25.061127.6529@cs.mcgill.ca>
Organization: Olekshy Hoover & Associates Ltd., Edmonton, Alberta, Canada.
Lines: 20
In-Reply-To: Message <1991Feb25.061127.6529@cs.mcgill.ca> dated 25 Feb 91 06:11:27 GMT

In message <1991Feb25.061127.6529@cs.mcgill.ca>, utility@cs.mcgill.ca
Ronald BODKIN writes:
>
> [...]                                           but I was hoping that
> someone could suggest how one might allow the safe execution of lisp
> forms from within a signal handler.  Maybe I should email David Betz?

The *only* thing you can safely do in a signal handler, in c, and be
portable, is set a flag, and perhaps only a flag of type sigatomic_t.
This has been discussed at length in comp.lang.c, and there's nothing
David can do about it.

Fortunately, the flag setting technique works nicely with the evalhook
technique.

--
Yours etc., Tony Olekshy.	Internet: tony%oha@CS.UAlberta.CA
				  BITNET: tony%oha.uucp@UALTAMTS.BITNET
				    uucp: alberta!oha!tony
There is nothing more dreadful than imagination without taste.  --Goethe