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: 1997/07/02
Message-ID: <dewar.867888791@merv>#1/1
X-Deja-AN: 254205385
References: <97061611223212@psavax.pwfl.com>
 <gwinn-1706972242370001@dh5055155.res.ray.com> <dewar.866846641@merv>
 <5olje1$oov$1@gonzo.sun3.iaf.nl>
Organization: New York University
Newsgroups: comp.lang.ada
Date: 1997-07-02T00:00:00+00:00
List-Id: <comp.lang.ada>


iGeert said

<<IMO it is an obvious bug to do memory I/O to such a device using
an unconstrained array type without any representation clauses or
whatever. From such a device you should only read data with 'Size=32
and pragma Atomic applied:
   type Bits is array (0 .. 31) of Boolean;
   pragma Pack (Bits);
   pragma Atomic (Bits);
   for Bits'Size use 32;

Certainly I would not expect any compiler to read elements from
Bits using byte reads in this case. For larger bit arrays than 32
bits I would choose to map the entire device memory using a
constrained array of Bits, like:
   type Word_Index is range 0 .. 2**32 - 1;
   type Words is array (Word_Index range <>) of Bits;
   subtype Video_Words is array (0 .. 16#3FFF#);
>>



Of course the code had the pragma Pack. I thought that was obvious
from context.

The Size clause is totally redundant, people often put it in for
good measure, but it has no effect, except to confirm the choice
that the compiler is making anyway.

The pramga Atomic is completely irrelevant, it affects only the semantics
of reads and writes to the entire array, not to its components.

<<Certainly I would not expect any compiler to read elements from
Bits using byte reads in this case>>

Yes, well the programmers in this case agreed with you. Unfortunately
you have absolutely NO basis for this expectation whatsoever. If you
think otherwise, try to prove it from the RM.

On some machines, the word read will be faster than the byte read

On some machines the byte read will be faster than the word read
(the 386SX is an example)

Om most machines, the speeds will be the same

For the first case, one hopes, but cannot be sure, that the compiler
will do word reads.

For the second case, one hopes, but cannot be sure, that the compiler
will do byte reads.

For the third case, one should have no hopes or expectations whatever!

It is these unjustified "Certainly I would not expect" intutions that
have no basis in the language definitoin that end up causing trouble!