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=-1.3 required=5.0 tests=BAYES_00,INVALID_MSGID
	autolearn=no autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,c78177ec2e61f4ac
X-Google-Attributes: gid103376,public
From: dewar@merv.cs.nyu.edu (Robert Dewar)
Subject: Re: ada and robots
Date: 1997/06/23
Message-ID: <dewar.867078319@merv>#1/1
X-Deja-AN: 252082974
References: <97061611223212@psavax.pwfl.com>
 <gwinn-1706972242370001@dh5055155.res.ray.com> <dewar.866846641@merv>
 <5ollh0$el9$1@news.nyu.edu>
Organization: New York University
Newsgroups: comp.lang.ada
Date: 1997-06-23T00:00:00+00:00
List-Id: <comp.lang.ada>


Richard said

<<This is perhaps a case where the casts in C are more portable.

If you write

        * ((volatile long *) some_address) |= some_expression;

it would be quite peculiar (and perhaps even be considered a bug) if the
compiler didn't respect the "long" type and generate the correspondingly
sized load and store instructions.
>>

I don't see this claim at all. The corresponding Ada code would be
identical semantically to the C code, and the same statement ("quite
peculiar .. perhaps even be considered a bug") would apply to the Ada.

But that was not the example, the example was where a single bit was flipped.

If we write   x &= 1;

in C where x is a volatile variable, there is nothing in the C language
that indicates whether we get

word load/store
byte load/store
bit set instruction (quite reasonable to expect on x86)


I think that the semantics of Ada and C are identical here. The advantage
of Ada is that it is a little more clear that you are playing low level
games -- in particular the quoted C code, in which the volatility is buried
in low level implementation, is undesirable.