comp.lang.ada
 help / color / mirror / Atom feed
From: john@assen.demon.co.uk (John McCabe)
Subject: Re: Ada Core Technologies and Ada95 Standards
Date: 1996/05/08
Date: 1996-05-08T00:00:00+00:00	[thread overview]
Message-ID: <831590385.6539.0@assen.demon.co.uk> (raw)
In-Reply-To: dewar.831423143@schonberg


dewar@cs.nyu.edu (Robert Dewar) wrote:

>John published an example of a supposed bug report. But it concerned
>pragma Shared, and pragma Shared only has non-null semantics in the
>presence of tasking, but there was no task in the program. So this
>is clearly not a bug report.

Robert, this topic has been discussed before by a number of people in
this group, and by email. I have also read your report on "Shared
Variables and Ada 9X Issues". It is quite clear from what is written
there that you believe the shared variables section of RM83 to be
extremely unclear, so please do not use the grammar of that section as
an argument defending a duff implementation, especially as you were so
involved in the effort to clear this mess up in Ada 9X.

>My guess is that you were assuming that pragma Shared should someone
>behave reasonably with respect to undefined outside accesses to something
>with an address clause. OK, that is a reasonable assumption, 

I would say that's more than a reasonable assumption given that (as I
stated towards the bottom of that message) the program works _exactly_
as I had expected if the address clause is left out. I cannot think of
any defence for an implementation to do this except that it is a bug -
hence this clearly _is_ a bug report.

>but there is no requirement to this effect in the RM.

True - but see above.

<.. snip ..>

>So you can't appeal to the RM here -- you can talk about what would be
>the most useful implementation approach. Certainly GNAT would not ignore
>the pragma in this situation, even though it is formally allowed to.

I imagine most compilers would not ignore the pragma in this
situation. I know, for example, that Tartan's compiler for
MIL-STD-1750A doesn't ignore the pragma. It is the consistency of the
implementation that is at fault here - why ignore the pragma when the
object is address claused (suggesting outside influences - possibly a
"phantom task"?) but take full account of it when no address clause is
given?

Anyway, this is getting a bit off-topic, it was only meant as an
example of a bug report that I had handy at the time. If you want to
dicuss this particular example further, please contact me via email
but I think it's been thrashed to death already.

Best Regards
John McCabe <john@assen.demon.co.uk>





  reply	other threads:[~1996-05-08  0:00 UTC|newest]

Thread overview: 106+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
1996-03-25  0:00 Ada Core Technologies and Ada95 Standards Kenneth Mays
1996-03-25  0:00 ` Robert Dewar
1996-03-28  0:00   ` John McCabe
1996-03-28  0:00     ` Robert Dewar
1996-03-29  0:00       ` John McCabe
1996-03-29  0:00         ` Robert Dewar
1996-04-01  0:00           ` Ken Garlington
1996-04-01  0:00             ` Robert Dewar
1996-04-02  0:00               ` Ken Garlington
1996-04-02  0:00                 ` John McCabe
1996-04-02  0:00                   ` Robert Dewar
1996-04-03  0:00                     ` Ken Garlington
1996-04-04  0:00                       ` Robert Dewar
1996-04-04  0:00                         ` Ken Garlington
1996-04-05  0:00                           ` Robert Dewar
1996-04-10  0:00                             ` Ken Garlington
1996-04-02  0:00                   ` Robert A Duff
1996-04-10  0:00                 ` Robert Dewar
1996-04-10  0:00                   ` Robert Dewar
1996-04-12  0:00                     ` Philip Brashear
1996-04-12  0:00                       ` Robert Dewar
1996-04-15  0:00                     ` Tiring Arguments Around (not about) Two Questions Ken Garlington
1996-04-15  0:00                       ` Gary McKee
1996-04-16  0:00                         ` Ken Garlington
1996-04-17  0:00                       ` Kenneth Almquist
1996-04-18  0:00                     ` Ada Core Technologies and Ada95 Standards John McCabe
1996-04-19  0:00                       ` Robert Dewar
1996-04-22  0:00                         ` John McCabe
1996-04-23  0:00                           ` Ken Garlington
1996-04-24  0:00                             ` Robert Dewar
1996-04-26  0:00                               ` Ken Garlington
1996-04-24  0:00                             ` John McCabe
1996-04-24  0:00                               ` Robert Dewar
1996-04-26  0:00                                 ` John McCabe
1996-04-26  0:00                                 ` John McCabe
1996-04-26  0:00                                 ` Ken Garlington
1996-04-25  0:00                               ` Ken Garlington
1996-04-24  0:00                           ` Robert Dewar
1996-04-26  0:00                             ` Ken Garlington
1996-04-27  0:00                               ` Robert Dewar
1996-04-22  0:00                         ` Ken Garlington
1996-04-15  0:00                   ` Ken Garlington
1996-04-16  0:00                     ` Robert Dewar
1996-04-16  0:00                       ` Ken Garlington
1996-04-16  0:00                         ` Robert Dewar
1996-04-02  0:00               ` John McCabe
1996-04-02  0:00             ` John McCabe
1996-04-02  0:00               ` Robert A Duff
1996-04-16  0:00                 ` John McCabe
1996-04-16  0:00                   ` Robert Dewar
1996-04-22  0:00                     ` John McCabe
1996-04-23  0:00                       ` Ken Garlington
1996-04-24  0:00                         ` Robert Dewar
1996-04-26  0:00                           ` Ken Garlington
1996-04-27  0:00                             ` Robert Dewar
1996-04-29  0:00                               ` Cordes MJ
1996-04-29  0:00                                 ` Robert Dewar
1996-05-06  0:00                                   ` John McCabe
1996-05-06  0:00                                     ` Robert Dewar
1996-05-08  0:00                                       ` John McCabe [this message]
1996-05-08  0:00                                         ` TARTAN and TI Tom Robinson
1996-05-09  0:00                                           ` Arthur Evans Jr
     [not found]                                         ` <Dr46LG.2FF@world.std.com>
1996-05-09  0:00                                           ` Ada Core Technologies and Ada95 Standards John McCabe
1996-05-07  0:00                                     ` Mike Cordes
1996-05-07  0:00                                     ` Mike Cordes
1996-04-10  0:00             ` Robert Dewar
1996-04-15  0:00               ` Ken Garlington
1996-04-16  0:00                 ` Robert Dewar
1996-04-16  0:00                   ` Ken Garlington
1996-04-16  0:00                     ` Robert Dewar
1996-04-18  0:00                       ` Ken Garlington
1996-03-31  0:00         ` Geert Bosch
1996-04-01  0:00           ` Robert Dewar
1996-04-01  0:00             ` Mike Young
1996-04-03  0:00               ` Robert Dewar
1996-03-29  0:00   ` Applet Magic works great, sort of Vince Del Vecchio
1996-03-29  0:00   ` Ada Core Technologies and Ada95 Standards steved
1996-03-29  0:00     ` Applet Magic works great, sort of Bob Crispen
1996-04-03  0:00   ` Ada Core Technologies and Ada95 Standards Ken Garlington
1996-04-04  0:00     ` Robert Dewar
1996-04-04  0:00       ` John McCabe
1996-04-05  0:00         ` Robert Dewar
1996-04-06  0:00           ` Ada validation is virtually worthless Raj Thomas
1996-04-06  0:00             ` Robert Dewar
1996-04-08  0:00               ` Arthur Evans Jr
1996-04-07  0:00           ` Ada Core Technologies and Ada95 Standards John McCabe
1996-04-03  0:00   ` Robert I. Eachus
1996-04-05  0:00   ` Robert I. Eachus
1996-04-10  0:00     ` Cordes MJ
1996-04-10  0:00       ` Robert Dewar
1996-04-15  0:00         ` Ken Garlington
1996-04-16  0:00           ` Robert Dewar
1996-04-16  0:00             ` Ken Garlington
1996-04-16  0:00               ` Robert Dewar
1996-04-11  0:00   ` Robert I. Eachus
1996-04-11  0:00   ` Robert I. Eachus
1996-04-19  0:00   ` Laurent Guerby
1996-04-25  0:00   ` Tiring Arguments Around (not about) Two Questions [VERY LONG] Laurent Guerby
1996-04-26  0:00   ` Ken Garlington
1996-04-29  0:00     ` Philip Brashear
  -- strict thread matches above, loose matches on Subject: below --
1996-04-01  0:00 Ada Core Technologies and Ada95 Standards Simon Johnston
1996-04-24  0:00 W. Wesley Groleau (Wes)
1996-04-25  0:00 ` Robert Dewar
1996-04-25  0:00   ` Theodore E. Dennison
1996-04-25  0:00 ` Ken Garlington
1996-04-29  0:00 Marin David Condic, 407.796.8997, M/S 731-93
replies disabled

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