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-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,d9e62a9b5f6f9797
X-Google-Attributes: gid103376,public
X-Google-ArrivalTime: 2002-01-25 12:37:10 PST
Path: archiver1.google.com!postnews1.google.com!not-for-mail
From: dewar@gnat.com (Robert Dewar)
Newsgroups: comp.lang.ada
Subject: Re: boolean array
Date: 25 Jan 2002 12:37:10 -0800
Organization: http://groups.google.com/
Message-ID: <5ee5b646.0201251237.766bac75@posting.google.com>
References: <mailman.1011938222.12962.comp.lang.ada@ada.eu.org>
NNTP-Posting-Host: 205.232.38.244
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
X-Trace: posting.google.com 1011991030 2722 127.0.0.1 (25 Jan 2002 20:37:10
 GMT)
X-Complaints-To: groups-abuse@google.com
NNTP-Posting-Date: 25 Jan 2002 20:37:10 GMT
Xref: archiver1.google.com comp.lang.ada:19312
Date: 2002-01-25T20:37:10+00:00
List-Id: <comp.lang.ada>

Christoph Grein <christoph.grein@eurocopter.com> wrote in message news:<mailman.1011938222.12962.comp.lang.ada@ada.eu.org>...
> Foo : array (some_range) of Boolean;
>     ...
>     Any_False_Present : constant Boolean := Foo /= (some_range => True);
> 
> But faster/preferred ??

If the array is packed, this should be fast. Here using the
latest version of GNAT, is a source program:

  procedure K is
    type Rng is range 1 .. 8;
    type a is array (Rng) of Boolean;
    pragma Pack (a);
    va : a := (1 => False, others => True);

  begin
     if va = (Rng => True) then 
        null;
     end if;
  end;

And here is the expanded source program:

  procedure k is
     type k__rng is range 1 .. 8;
     type k__a is array (1 .. 8) of boolean;
     pragma pack (k__a);
     va : k__a := k__T6b!(16#FE#);
  begin
     if va = k__T10b!(16#FF#) then
        null;
     end if;
     return;
  end k;

The ! here is unchecked conversion, and the implementation
type of va is an 8 bit unsigned number, so as you can see
this reduces the cost to a single comparison instruction.

For those who need to see the asm to be convinced, here 
it is:

	.file	"k.adb"
gcc2_compiled.:
___gnu_compiled_ada:
.text
	.align 2
.globl __ada_k
__ada_k:
	pushl %ebp
	movl %esp,%ebp
	subl $4,%esp
	movb $254,-1(%ebp)
	cmpb $255,-1(%ebp)
	jne L2
L2:
	jmp L1
	.align 2,0x90
L1:
	leave
	ret