comp.lang.ada
 help / color / mirror / Atom feed
* Alsys Executable
@ 2001-03-13 18:06 u535034179
  2001-03-14 21:22 ` Stephen Leake
  0 siblings, 1 reply; 4+ messages in thread
From: u535034179 @ 2001-03-13 18:06 UTC (permalink / raw)


Goal: Generate identical output from identical source code using Alsys
(Aonix) Ada Compiler v2a:  Host: Sparc, Target: M68K

Problem:  Resultant S-Record executables differ.  After disassembling
the S-Record output, there appears to be a lot of debug information 
which is the source of these differences.

Questions:
1. Is there a way to produce the executable w/o any extraneous 
information? I have not found this in any documentation.  Compiler is 
no longer supported by Aonix.
2. Is there a better way to purse this goal?

Thanks in advance,
Chris

 



-- 
Sent by tlutton from  aol subdomain of com
This is a spam protected message. Please answer with reference header.
Posted via http://www.usenet-replayer.com/cgi/content/new



^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Alsys Executable
       [not found] <20010313183703.183FD191EE@ada.eu.org>
@ 2001-03-13 19:02 ` Mike Brenner
  0 siblings, 0 replies; 4+ messages in thread
From: Mike Brenner @ 2001-03-13 19:02 UTC (permalink / raw)
  To: comp.lang.ada

> Goal: Generate identical output from identical source code using Alsys
> (Aonix) Ada Compiler v2a:  Host: Sparc, Target: M68K
>
> Problem:  Resultant S-Record executables differ.  After disassembling
> the S-Record output, there appears to be a lot of debug information
> which is the source of these differences.
> 
> Questions:
> 1. Is there a way to produce the executable w/o any extraneous
> information? I have not found this in any documentation.  Compiler is
> no longer supported by Aonix.
> 2. Is there a better way to purse this goal?


I do not understand the goal, but if the problem is removing debug
information from an Alsys compiler output, here are a couple of things
to try.

On the compiler, try turning off checks. Try KEEP DEBUG NO. Try OPTIONS
GENERICS STUBS. Try DISPLAY ASSEMBLY NONE. Try KEEP TREE NO.

On the binder, try KEEP DEBUG NO. Try DISPLAY DATA NO. Try STACK HISTORY
NO. 



Of these, the ones that look the most like debug information are, in
priority order:

	STACK HISTORY NO and KEEP DEBUG NO on the binder
and
	KEEP DEBUG NO and KEEP TREE NO

Of course, some of these are kept in the LIBRARY versus the executable,
but that's okay becuase I did not understand your goal. If you truly
meant only the executable, then start with:

	STACK HISTORY NO in the binder

Mike Brenner


P.S. Here are the defaults for the compiler.

1. default.compile (
-- source  => ?,
-- library => ?,
   options => (
            -- annotate    => *,
            -- errors      => *,
            -- checks      => none,
               checks      => all,
            -- checks      => stack,
            -- generics    => inline,
               generics    => stubs,
            -- memory      => 500,
               memory      => 1000,
               tasking     => no),
   display => (
               output      => screen,
            -- output      => automatic,
            -- output      => none,
               warning     => yes,
               text        => no,
            -- show        => banner,
            -- show        => recap,
               show        => all,
            -- show        => none,
               detail      => yes,
            -- assembly    => all
            -- assembly    => code),
            -- assembly    => map),
               assembly    => none),
   improve => (
               calls       => normal,
            -- calls       => inlined,
               reduction   => none,
            -- reduction   => partial,
            -- reduction   => extensive,
               expressions => none),
            -- expressions => partial),
            -- expressions => extensive),
  allocation => (stack=>252),  -- The size of fixed_strings
  keep       => (
               copy        => no,
               debug       => no,
               tree        => no))
            -- copy        => yes,
            -- debug       => yes,
            -- tree        => yes))


default.bind    (
   program    => test,
-- library    => ?
   options    => (
                   level       => link,
                -- level       => check,
                -- level       => bind,
                -- execution   => extended,
                   execution   => protected,
                -- float       => automatic,
                   float       => hardware,
                -- float       => software,
                   mathlib     => I387,
                -- object      => automatic,
                   object      => g:.,
                -- uncalled    => remove,
                   uncalled    => keep,
                   timer       => normal,
                -- timer       => fast,
                   slice       => no),

   stack      => (
                -- main        => 96
                   main        => 127,
                   task        => 1,
                   history     => yes), -- Traceback in the .CUI file

   heap       => (
                -- size        => 64,
                   size        => 128,
                -- increment   => 64),
                   increment   => 128),


--
-- NOTE: the "modules" below are needed for BIOS calls
--
   interface  => (
                -- search      => *,
                -- directives  => *,
                   modules     =>
   "c:\obj\vectora.obj c:\obj\getphys.obj c:\obj\mapphys.obj"),

   display    => (
                   data  =>    none),
                -- data  =>    all), -- link map

   keep       => (
                   debug =>    no));

workbench;
default.xref (output=> screen );
quit; -- workbench;





^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Alsys Executable
  2001-03-13 18:06 u535034179
@ 2001-03-14 21:22 ` Stephen Leake
  2001-03-15  2:15   ` Keith Thompson
  0 siblings, 1 reply; 4+ messages in thread
From: Stephen Leake @ 2001-03-14 21:22 UTC (permalink / raw)


u535034179@spawnkill.ip-mobilphone.net writes:

> Goal: Generate identical output from identical source code using Alsys
> (Aonix) Ada Compiler v2a:  Host: Sparc, Target: M68K

Identical to what? You've only specified one thing here.

-- 
-- Stephe



^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Alsys Executable
  2001-03-14 21:22 ` Stephen Leake
@ 2001-03-15  2:15   ` Keith Thompson
  0 siblings, 0 replies; 4+ messages in thread
From: Keith Thompson @ 2001-03-15  2:15 UTC (permalink / raw)


Stephen Leake <stephen.a.leake.1@gsfc.nasa.gov> writes:
> u535034179@spawnkill.ip-mobilphone.net writes:
> > Goal: Generate identical output from identical source code using Alsys
> > (Aonix) Ada Compiler v2a:  Host: Sparc, Target: M68K
> 
> Identical to what? You've only specified one thing here.

I think the problem is that two compilations of the same source
generate non-identical (but presumably functionally equivalent)
executables.  Probably some of the data in the executable either
depends on things that change from one compilation to another
(timestamps, etc) or isn't used and is left in an undefined state.

I don't have a solution -- except perhaps finding another solution to
the original problem that doesn't require identical executables.

-- 
Keith Thompson (The_Other_Keith) kst@cts.com  <http://www.ghoti.net/~kst>
San Diego Supercomputer Center           <*>  <http://www.sdsc.edu/~kst>
MAKE MONEY FAST!!  DON'T FEED IT!!



^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2001-03-15  2:15 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
     [not found] <20010313183703.183FD191EE@ada.eu.org>
2001-03-13 19:02 ` Alsys Executable Mike Brenner
2001-03-13 18:06 u535034179
2001-03-14 21:22 ` Stephen Leake
2001-03-15  2:15   ` Keith Thompson

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox