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=0.2 required=5.0 tests=BAYES_00,INVALID_MSGID,
	REPLYTO_WITHOUT_TO_CC autolearn=no autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,53a810f13e8ba40
X-Google-Attributes: gid103376,public
From: Mats Weber <Mats.Weber@elca-matrix.ch>
Subject: Re: Byte/Bit Twiddling in Ada
Date: 1997/02/14
Message-ID: <33048844.7AEB@elca-matrix.ch>#1/1
X-Deja-AN: 218962768
References: <5dvfnn$olj@neocad.xilinx.com>
Content-Type: text/plain; charset=us-ascii
Organization: ELCA Matrix SA
Mime-Version: 1.0
Reply-To: Mats.Weber@elca-matrix.ch
Newsgroups: comp.lang.ada
X-Mailer: Mozilla 3.01 (Macintosh; I; PPC)
Date: 1997-02-14T00:00:00+00:00
List-Id: <comp.lang.ada>


> Is there any portable way to do byte and bit manipulation in Ada, on a
> long word?  I found no mention of this topic in the FAQ.

longword_bits : constant := 32;

type longword is array (0 .. longword_bits - 1) of boolean;
pramga pack(longword);
for longword'size use longword_bits;

then you can do any bit-twiddling on variables of type Longword, e.g.

A, B : Longword;

A(0 .. 15) := A(16 .. 31);  -- copy low-order 16 bits 
                            -- into high order 16 bits

A(0 .. 30) := A(1 .. 31);  -- right shift
A := A(31) & A(0 .. 30);   -- rotate left

You can use Unchecked_Conversion if you want to manipulate floating
point values at the bit level (at your own risk). In this case, set
Longword_Bits = <your floating point type>'Size;

(Robert, what kind of code would GNAT generate for these instructions ?
Does the optimizer recognize the CPU's instruction set ?)

> My friend is trying to write some code that does byte swapping,
> presumably to handle conversions to and from IEEE number formats.

Another possibility is using representation clauses:

type Mantissa is range -2 ** 22 .. 2 ** 22 - 1;
type Exponent is range -2 ** 8 .. 2 ** 8 - 1;

type Float_Rec is
   record
      Mant : Mantissa;
      Expo : Exponent;
   end record;

pragma Pack(Float_Rec);
for Float_Rec'Size use 32;

for Float_Rec use 
   record at mod 4;  -- if you want them aligned
      Mant at 0 range 0 .. 22;
      Expo at 0 range 23 .. 31;
   end record;

Note: I did this without looking at the reference of any floating point
type, and I did not check the code with a compiler. But it should work
once adapted.