From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=-1.9 required=5.0 tests=BAYES_00 autolearn=ham
	autolearn_force=no version=3.4.4
X-Google-Thread: 103376,b3f788f59498d3af
X-Google-Attributes: gid103376,public,usenet
X-Google-Language: ENGLISH,ASCII-7-bit
Path: 
 g2news1.google.com!news4.google.com!feeder1-2.proxad.net!proxad.net!feeder1-1.proxad.net!club-internet.fr!feedme-small.clubint.net!nuzba.szn.dk!news.jacob-sparre.dk!pnx.dk!not-for-mail
From: "Randy Brukardt" <randy@rrsoftware.com>
Newsgroups: comp.lang.ada
Subject: Re: Exceptions and out procedure arguments (using GNAT GPL)
Date: Fri, 15 Jun 2007 22:08:44 -0500
Organization: Jacob's private Usenet server
Message-ID: <f4vk3j$mu3$1@jacob-sparre.dk>
References: <79c673pq5htg508nkoi935n3udqg5ps7r8@4ax.com>
NNTP-Posting-Host: static-69-95-181-76.mad.choiceone.net
X-Trace: jacob-sparre.dk 1181963187 23491 69.95.181.76 (16 Jun 2007 03:06:27
 GMT)
X-Complaints-To: news@jacob-sparre.dk
NNTP-Posting-Date: Sat, 16 Jun 2007 03:06:27 +0000 (UTC)
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2800.1807
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1896
Xref: g2news1.google.com comp.lang.ada:16205
Date: 2007-06-15T22:08:44-05:00
List-Id: <comp.lang.ada>

"Fionn Mac Cumhaill" <invisible@hiding.from.spam> wrote in message
news:79c673pq5htg508nkoi935n3udqg5ps7r8@4ax.com...
...
> Will X get a value of 999? I.e., is an Ada compiler free to optimize
> the code such that the assignment to the out variable O is moved to
> some place after the point where the exception is raised?

My gut feeling (not one backed up by checking the rules in detail) is that
it depends on how the parameters are passed. Integers are a by-copy type, so
a compiler that modified X would be wrong: the copy back into X occurs on
the return from the call, and when the exception is raised that copy back
will not be executed.

Conversely, if the parameter is a by-reference type, then the assignment
should be reflected. (But I'd have to read the 11.6 rules ten times to be
sure, and it probably would depend on the actual code in the subprogram.)
Surely, Janus/Ada would make the assignment in that case.

If the language does not specify the parameter passing mechanism (as for
untagged records and arrays), then either thing could happen and you should
depend on nothing.

Surely the safest rule is to never depend on the parameter being changed,
but either thing can happen.

                           Randy.