'address on a subprogram parameter

'address on a subprogram parameter

[Vincent Marciante]

I think that I used to know this but currently am not sure:

Assuming the a value is passed by reference, is 'Address of a parameter 
always the address of the object that is passed as in:

...

  type Big_Record is ...

  Object : Big_Record;

  procedure Proc (Param :  in out Big_Record) is
  begin
    if  Param'Address /= Object'Address  then
         raise Constraint_Error;
    end if;     
  end;

begin
  Proc(Object);
  ...


Also, does parameter mode (in, in out, out) make any difference? 

Re: 'address on a subprogram parameter

[AdaMagica]

I'm not a compiler builder, but pass by referene has always meant, you access the parameter object directly. And of course,  the parameter mode does not (and must not) have any influence on the parameter passing method.
You can force pass by ref by adding access to the parameter.
In most cases, passing method is of no interest, compiler makes the correct choice. RM is silent on most types about this.
Tagged types and limited types are always PbR.

Re: 'address on a subprogram parameter

[Jeffrey R. Carter]

On 4/20/21 5:52 PM, AdaMagica wrote:
> You can force pass by ref by adding access to the parameter.

I hope you meant to say "aliased", not "access". "Access" does not force pass by 
reference; it forces pass by copy because access types are always passed by copy.

-- 
Jeff Carter
"I'm particularly glad that these lovely children were
here today to hear that speech. Not only was it authentic
frontier gibberish, it expressed a courage little seen
in this day and age."
Blazing Saddles
88

Re: 'address on a subprogram parameter

[AdaMagica]

Jeffrey R. Carter schrieb am Dienstag, 20. April 2021 um 19:32:08 UTC+2:
> On 4/20/21 5:52 PM, AdaMagica wrote: 
> > You can force pass by ref by adding access to the parameter.
> I hope you meant to say "aliased", not "access". "Access" does not force pass by 
> reference; it forces pass by copy because access types are always passed by copy. 

Of course you're right. Sorry for the confusion.
The access value is passed by copy, if you then dereference, you get at the actual object.
I reall meant alias.

Re: 'address on a subprogram parameter

[Niklas Holsti]

On 2021-04-20 18:52, AdaMagica wrote:

> You can force pass by ref by adding access to the parameter.

Or adding "aliased", which does not force one to use 'Access in the call.

Re: 'address on a subprogram parameter

[J-P. Rosen]

Le 20/04/2021 à 17:10, Vincent Marciante a écrit :
> I think that I used to know this but currently am not sure:
> 
> Assuming the a value is passed by reference, is 'Address of a parameter
> always the address of the object that is passed as in:
> 
> ...
> 
>    type Big_Record is ...
> 
>    Object : Big_Record;
> 
>    procedure Proc (Param :  in out Big_Record) is
>    begin
>      if  Param'Address /= Object'Address  then
>           raise Constraint_Error;
>      end if;
>    end;
> 
> begin
>    Proc(Object);
>    ...
> 
As usual, the real question is "why do you need that?"

If you want to check whether a parameter IS a certain variable, use 
Param'Has_Same_Storage(Object). If you want to check whether variables 
have a part in common, use Param'Overlaps_Storage (Object).
See 13.3 (73.1/3 .. 73.10/3)


-- 
J-P. Rosen
Adalog
2 rue du Docteur Lombard, 92441 Issy-les-Moulineaux CEDEX
Tel: +33 1 45 29 21 52
https://www.adalog.fr

Re: 'address on a subprogram parameter

[Vincent Marciante]

On Tuesday, April 20, 2021 at 12:35:07 PM UTC-4, J-P. Rosen wrote:
> Le 20/04/2021 à 17:10, Vincent Marciante a écrit : 
> > I think that I used to know this but currently am not sure: 
> > 
> > Assuming the a value is passed by reference, is 'Address of a parameter 
> > always the address of the object that is passed as in: 
> > 
> > ... 
> > 
> > type Big_Record is ... 
> > 
> > Object : Big_Record; 
> > 
> > procedure Proc (Param : in out Big_Record) is 
> > begin 
> > if Param'Address /= Object'Address then 
> > raise Constraint_Error; 
> > end if; 
> > end; 
> > 
> > begin 
> > Proc(Object); 
> > ... 
> >
> As usual, the real question is "why do you need that?" 
> 
> If you want to check whether a parameter IS a certain variable, use 
> Param'Has_Same_Storage(Object). If you want to check whether variables 
> have a part in common, use Param'Overlaps_Storage (Object). 
> See 13.3 (73.1/3 .. 73.10/3) 
> 
> 
> -- 
> J-P. Rosen 
> Adalog 
> 2 rue du Docteur Lombard, 92441 Issy-les-Moulineaux CEDEX 
> Tel: +33 1 45 29 21 52 
> https://www.adalog.fr

I actually do not want to test that: I just wrote that as an attempt to
make explicit want I was asking and what I did not expect (exception).
I am really working with old code where I am moving some logic out 
from within a package body that contains the "address'd" variable to
an unassociated package that defines the "Big_Record" type and I
wanted to be sure that passing it as a parameter would (always) have
the same behavior with the variable being passed as a parameter.
 

Re: 'address on a subprogram parameter

[Randy Brukardt]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1103 bytes --]

"Vincent Marciante" <vincent.marciante@l3harris.com> wrote in message 
news:13f272e5-14a3-45cb-abf2-3d9f5a98f715n@googlegroups.com...
On Tuesday, April 20, 2021 at 12:35:07 PM UTC-4, J-P. Rosen wrote:
> Le 20/04/2021 à 17:10, Vincent Marciante a écrit :
> > I think that I used to know this but currently am not sure:
> >
> > Assuming the a value is passed by reference,

You can only assume that for by-reference types (see RM 6.3(4-9)).

Your example just used an unspecified record type. For any old record type 
(not in that list), the parameter passing mechanism is unspecified. Which 
means a compiler can pass by copy if it wants (it *probably* will pass by 
reference, but that's not the same as being required to do so). And if it 
does, your assumption doesn't hold.

For instance, for a small record like:
    type Screen_Position is
       X, Y : System.Byte; -- A Janus/Adaism, but I think you get the point.
    end record;
it might make more sense to pass by copy (especially if registers are used 
in the usual calling convention).

                    Randy. 

Re: 'address on a subprogram parameter

[Vincent Marciante]

On Tuesday, April 20, 2021 at 3:25:25 PM UTC-4, Randy Brukardt wrote:
> "Vincent Marciante" wrote in message 
> On Tuesday, April 20, 2021 at 12:35:07 PM UTC-4, J-P. Rosen wrote: 
> > Le 20/04/2021 à 17:10, Vincent Marciante a écrit : 
> > > I think that I used to know this but currently am not sure: 
> > > 
> > > Assuming the a value is passed by reference,
> You can only assume that for by-reference types (see RM 6.3(4-9)). 
> 
> Your example just used an unspecified record type. For any old record type 
> (not in that list), the parameter passing mechanism is unspecified. Which 
> means a compiler can pass by copy if it wants (it *probably* will pass by 
> reference, but that's not the same as being required to do so). And if it 
> does, your assumption doesn't hold. 
> 
> For instance, for a small record like: 
> type Screen_Position is 
> X, Y : System.Byte; -- A Janus/Adaism, but I think you get the point. 
> end record; 
> it might make more sense to pass by copy (especially if registers are used 
> in the usual calling convention). 
> 
> Randy.

Got it (and "no way" that the big record that is in the real code would be passed by copy
but I'll see if specifying it explicitly would not be disruptive).  Thanks.

Re: 'address on a subprogram parameter

[Randy Brukardt]

"Vincent Marciante" <vincent.marciante@l3harris.com> wrote in message 
news:07a090e1-85b4-4ebb-a093-dcda57b56579n@googlegroups.com...
>Got it (and "no way" that the big record that is in the real code would be 
>passed by copy
>but I'll see if specifying it explicitly would not be disruptive).  Thanks.

Careful about assumptions, if you want the code to be portable. At one 
point, Janus/Ada would pass pretty anything by copy to parameters of certain 
generic types. (We got rid of that in the Ada 95 version because of the need 
to deal with "aliased" components, but the cost is that one needs to use a 
thunk to access anything of most generic formal types in a generic body. 
[Janus/Ada still uses universal generic sharing.])

                                   Randy. 

Re: 'address on a subprogram parameter

[Niklas Holsti]

On 2021-04-24 4:25, Randy Brukardt wrote:
> "Vincent Marciante" <vincent.marciante@l3harris.com> wrote in message
> news:07a090e1-85b4-4ebb-a093-dcda57b56579n@googlegroups.com...
>> Got it (and "no way" that the big record that is in the real code would be
>> passed by copy
>> but I'll see if specifying it explicitly would not be disruptive).  Thanks.
> 
> Careful about assumptions, if you want the code to be portable.


Seconded.

Long ago I was working with early "public" GNAT versions ("gnatp") on a 
certain program. After we upgraded to the next version of gnatp (I think 
around version 3.12) our program started failing with Storage_Error. 
Turned out that the GNAT developers had decided that this gnatp version 
would pass all record-type parameters by value, "to be more like C" 
(IIRC Robert Dewar was said to be the origin of that idea, but that may 
be a calumny). And our program had some large record types.

We ended up shipping the program with the advice to use a 10 MiB stack 
allocation -- but it worked (it did not have any real-time requirements).

As a positive, the value-passing gnatp version made us discover a couple 
of unintended parameter-vs-global aliasing issues that had been hidden 
when all records were passed by reference.

The very next gnatp version returned to pass-by-reference for 
record-type objects. It seems there had been some negative user 
feed-back :-)


> At one point, Janus/Ada would pass pretty anything by copy to 
> parameters of certain generic types. (We got rid of that in the Ada
> 95 version because of the need to deal with "aliased" components, but
> the cost is that one needs to use a thunk to access anything of most
> generic formal types in a generic body. [Janus/Ada still uses
> universal generic sharing.])
>                                     Randy.

Re: 'address on a subprogram parameter

[Shark8]

On Tuesday, April 20, 2021 at 11:18:24 AM UTC-6, Vincent Marciante wrote:
> I actually do not want to test that: I just wrote that as an attempt to 
> make explicit want I was asking and what I did not expect (exception). 
> I am really working with old code where I am moving some logic out 
> from within a package body that contains the "address'd" variable to 
> an unassociated package that defines the "Big_Record" type and I 
> wanted to be sure that passing it as a parameter would (always) have 
> the same behavior with the variable being passed as a parameter.
Why not use genreics?

Generic
  Object : in out Big_Record;
Procedure Do_Something;

Remember, also, that you can supply subprograms in the formal parameters.

Re: 'address on a subprogram parameter

[Vincent Marciante]

On Wednesday, April 21, 2021 at 4:17:43 PM UTC-4, Shark8 wrote:
> On Tuesday, April 20, 2021 at 11:18:24 AM UTC-6, Vincent Marciante wrote: 
> > I actually do not want to test that: I just wrote that as an attempt to 
> > make explicit want I was asking and what I did not expect (exception). 
> > I am really working with old code where I am moving some logic out 
> > from within a package body that contains the "address'd" variable to 
> > an unassociated package that defines the "Big_Record" type and I 
> > wanted to be sure that passing it as a parameter would (always) have 
> > the same behavior with the variable being passed as a parameter.
> Why not use genreics? 
> 
> Generic 
> Object : in out Big_Record; 
> Procedure Do_Something; 
> 
> Remember, also, that you can supply subprograms in the formal parameters.

Using a method that ensure that the parameter will be passed by reference 
is all that I need.  Generics would just add one extra step - the instantiation.