comp.lang.ada
 help / color / mirror / Atom feed
* "limited with" with gnat GPL
@ 2006-10-20 17:46 evangeli
  2006-10-20 22:35 ` Georg Bauhaus
  0 siblings, 1 reply; 7+ messages in thread
From: evangeli @ 2006-10-20 17:46 UTC (permalink / raw)


Hello,
I am having problems in my ada application to get "limited with" work
on gnat.
Each time I do a "limited with" either the gnat crashs either I get
weird error messages such as :
pn-expressions.adb:30:03 "" is undefined
where line 30 is:
case Get_Type(E) is

So I would really appreciate if someone had some hints to have this
feature work, because I have many mutually recursive types that I must
currently put in the same package, which is not very pretty.

My configuration is the following one:
GNATLS GPL 2006 (20060522-34)
Suse 10.0 (i586)
Kernel 2.6.13

Thanks




^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: "limited with" with gnat GPL
  2006-10-20 17:46 "limited with" with gnat GPL evangeli
@ 2006-10-20 22:35 ` Georg Bauhaus
  2006-10-21 11:17   ` evangeli
  0 siblings, 1 reply; 7+ messages in thread
From: Georg Bauhaus @ 2006-10-20 22:35 UTC (permalink / raw)


On Fri, 2006-10-20 at 10:46 -0700, evangeli@cnam.fr wrote:
> Hello,
> I am having problems in my ada application to get "limited with" work
> on gnat.

Just to make sure there isn't a setup problem,
have you checked the PATH variable? Is there another
GNAT installed? Also, could you post an example that
fails, so others can check?




-- Georg 





^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: "limited with" with gnat GPL
  2006-10-20 22:35 ` Georg Bauhaus
@ 2006-10-21 11:17   ` evangeli
  2006-10-21 11:56     ` Georg Bauhaus
  0 siblings, 1 reply; 7+ messages in thread
From: evangeli @ 2006-10-21 11:17 UTC (permalink / raw)


> > Hello,
> > I am having problems in my ada application to get "limited with" work
> > on gnat.
>
> Just to make sure there isn't a setup problem,
> have you checked the PATH variable? Is there another
> GNAT installed? Also, could you post an example that
> fails, so others can check?
>
>
>
>
> -- Georg

doesn't seem to be a problem of PATH. I have only one gnat installed
and here is the result of some commands:

> which gnatmake
/usr/gnat/bin/gnatmake
> which gcc
/usr/gnat/bin/gcc
> which gnatls
/usr/gnat/bin/gnatls

here is a sample code that makes gnat crash

--==========
--  file pn-places.ads
limited with Pn.Arcs;  --  line X

with
  Generic_Array,
  Pn.Expressions,
  Pn.Mappings,
  Pn.Markings;

use
  Pn.Expressions,
  Pn.Mappings,
  Pn.Markings;

package Pn.Places is

   A : access Pn.Arcs.Arcs_List;  --  line Y
   ...
end Pn.Places;
--==========

gnat crashes and outputs this :

+===========================GNAT BUG
DETECTED==============================+
| GPL 2006 (20060522-34) (i686-pc-linux-gnu) Gigi abort, Code=320
| Error detected at pn-places.ads:551:4
...
+==========================================================================+

the line at 551 in pn-places.ads is the beginning of a tagged record
type :
   type Place_Record is tagged record
      Name : Ustring;
      D : Pn.Expressions.Dom;
      T : Place_Type;
      M0 : Pn.Mappings.Mapping;
      Capacity : Mult_Type;
      Redundant : Boolean;
      Data_Type : Marking_Type;
      Selected  : Boolean;
   end record;

if I remove lines X and Y from file pn-places.ads, the compilation
works fine.
I made an archive of the files reported by gnat :
http://deptinfo.cnam.fr/~evange_s/test-gnat.tgz
could someone please test it by compiling :
gnatmake pn-places.adb

thanks for help




^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: "limited with" with gnat GPL
  2006-10-21 11:17   ` evangeli
@ 2006-10-21 11:56     ` Georg Bauhaus
  2006-10-21 14:03       ` evangeli
  0 siblings, 1 reply; 7+ messages in thread
From: Georg Bauhaus @ 2006-10-21 11:56 UTC (permalink / raw)


evangeli@cnam.fr wrote:

> gnat crashes and outputs this :
> 
> +===========================GNAT BUG
> DETECTED==============================+
> | GPL 2006 (20060522-34) (i686-pc-linux-gnu) Gigi abort, Code=320
> | Error detected at pn-places.ads:551:4
> ...
> +==========================================================================+

This is an instance of a GNAT Bug Box. Your code might be fine,
but it triggers a compiler error. Gigi translates front end data
into data for the compiler middle end.



^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: "limited with" with gnat GPL
  2006-10-21 11:56     ` Georg Bauhaus
@ 2006-10-21 14:03       ` evangeli
  2006-10-22 11:44         ` Stephen Leake
  2006-10-22 16:29         ` Georg Bauhaus
  0 siblings, 2 replies; 7+ messages in thread
From: evangeli @ 2006-10-21 14:03 UTC (permalink / raw)


> evangeli@cnam.fr wrote:
>
> > gnat crashes and outputs this :
> >
> > +===========================GNAT BUG
> > DETECTED==============================+
> > | GPL 2006 (20060522-34) (i686-pc-linux-gnu) Gigi abort, Code=320
> > | Error detected at pn-places.ads:551:4
> > ...
> > +==========================================================================+
>
> This is an instance of a GNAT Bug Box. Your code might be fine,
> but it triggers a compiler error. Gigi translates front end data
> into data for the compiler middle end.

so the bad news seems to be that the problem comes from GNAT and that I
have to wait for a bug correction?
do you think that testing on another platform could produce other
results?




^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: "limited with" with gnat GPL
  2006-10-21 14:03       ` evangeli
@ 2006-10-22 11:44         ` Stephen Leake
  2006-10-22 16:29         ` Georg Bauhaus
  1 sibling, 0 replies; 7+ messages in thread
From: Stephen Leake @ 2006-10-22 11:44 UTC (permalink / raw)


"evangeli@cnam.fr" <evangeli@cnam.fr> writes:

>> evangeli@cnam.fr wrote:
>>
>> > gnat crashes and outputs this :
>> >
>> > +===========================GNAT BUG
>> > DETECTED==============================+
>> > | GPL 2006 (20060522-34) (i686-pc-linux-gnu) Gigi abort, Code=320
>> > | Error detected at pn-places.ads:551:4
>> > ...
>> > +==========================================================================+
>>
>> This is an instance of a GNAT Bug Box. Your code might be fine,
>> but it triggers a compiler error. Gigi translates front end data
>> into data for the compiler middle end.
>
> so the bad news seems to be that the problem comes from GNAT and that I
> have to wait for a bug correction?

You should submit a bug report, as the bug box says.

> do you think that testing on another platform could produce other
> results?

I have a more recent version of GNAT; post compilable code and I'll
test it.

-- 
-- Stephe



^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: "limited with" with gnat GPL
  2006-10-21 14:03       ` evangeli
  2006-10-22 11:44         ` Stephen Leake
@ 2006-10-22 16:29         ` Georg Bauhaus
  1 sibling, 0 replies; 7+ messages in thread
From: Georg Bauhaus @ 2006-10-22 16:29 UTC (permalink / raw)


evangeli@cnam.fr wrote:

>> This is an instance of a GNAT Bug Box....
> 
> so the bad news seems to be that the problem comes from GNAT and that I
> have to wait for a bug correction?
> do you think that testing on another platform could produce other
> results?
> 

One of the good things about GNAT is that it is easy to get or build
different versions if you cannot otherwise get a supported compiler.
 The Debian versions, for example,  come with some bug fixes
to the standard GNU distribution. GCC 4.2 is also fairly good.

Another option is, of course, to ask for support (which usually
either means that your project is academically or otherwise
interesting for AdaCore so they help you, or you buy a subscription.
Or try another Ada compiler if there is a budget.)



^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2006-10-22 16:29 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2006-10-20 17:46 "limited with" with gnat GPL evangeli
2006-10-20 22:35 ` Georg Bauhaus
2006-10-21 11:17   ` evangeli
2006-10-21 11:56     ` Georg Bauhaus
2006-10-21 14:03       ` evangeli
2006-10-22 11:44         ` Stephen Leake
2006-10-22 16:29         ` Georg Bauhaus

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