From: Jean-Marc Bourguet <jm@bourguet.org>
Subject: Re: GNAT equivalent for Green Hills functionality
Date: 25 Apr 2002 16:53:11 +0200
Date: 2002-04-25T16:53:11+02:00 [thread overview]
Message-ID: <3cc81861$1@news.cadence.com> (raw)
In-Reply-To: a87a922b.0204250638.1018e623@posting.google.com
michael.p.card@lmco.com (Michael P. Card) writes:
[Gren Hills provided 3 C functions]
> rts_task_init()- registers a C thread with Ada run-time so that after
> this call Ada will see it as if it were an Ada task. This way the C code
> can call Ada code that might make a protected object call, task
> rendezvous etc and teh C thread will be suspended and restarted properly
> just as if it were an Ada task
I seem to remember that there is something similar with gnat, but I've
not used it.
> adainit()- kicks off all Ada elaboration, initialization etc just as
> would happen at program start if Ada were the "main"
>
> adafinal()- finalizes all objects just as would occur at end-of-program
> if Ada were the "main"
These are defined in the standard, and I've used them with gnat.
> Does GNAT provide similar capability for cases where C++ is the "main"?
> Note that it may be that we will be using a C++ compiler other than gcc,
> I have heard that Ada/C++ integration is easy with GNAT if the C++ code
> is compiled with gcc.
From what I've understood, Ada/C++ integration is easy with GNAT,
whatever compiler is used for C++. It comes with the interface set up
for g++ but you can describe your C++ compiler. And obviously if you
are interfacing with C calling convention (using only extern "C" C++
functions), then the way of doing it is standard.
I suggest to contact ACT, your email address makes me doubt that you
should be using GNAT without support and from what I get, the quality
of their support is their main selling argument.
Yours,
--
Jean-Marc
prev parent reply other threads:[~2002-04-25 14:53 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-04-25 14:38 GNAT equivalent for Green Hills functionality Michael P. Card
2002-04-25 14:53 ` Jean-Marc Bourguet [this message]
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox