comp.lang.ada
 help / color / mirror / Atom feed
From: "Hartmut H. Schaefer" <Hartmut_Peg@compuserve.com>
Subject: Re: Ada compilers and RTOSs
Date: 1998/01/09
Date: 1998-01-09T00:00:00+00:00	[thread overview]
Message-ID: <01bd1d41$ded7b440$7a24e8c3@hup-pc> (raw)
In-Reply-To: 6939p7$dil$2@paperboy.ids.net




jkennedy <jkennedy@seacorp.com> schrieb im Beitrag
<6939p7$dil$2@paperboy.ids.net>...
> Hello,
>  
> I am working on a system that is using the Ready Systems Ada Real-Time 
> Executive (ARTX) RTOS with RTAda, and a Telesoft Telgen2 Ada compiler.
Our 
> system is currently hosted on a Vax with a 68K target. 
> 
> At this time we are trying to update our development environment,and we'd

> like to get off the Vax, and onto a Win NT based machine. However, the
ARTX 
> RTOS and the Telesoft compiler are no longer supported by their
respective 
> OEMs and the latest Ada 95 compilers for 68K are targeting Wind River's 
> VxWorks RTOS. 
> 
> I've been looking at Green Hill's AdaMULTI system for Tornado, but this
would 
> require a substantial re-code and we'd really like to minimize any
impacts to 
> our existing code. Has anyone "out there" either had or is having a
similar 
> problem ?
> 
> Any suggestions / inputs would be greatly appreciated.
>  						
> 						Thanks in advance,
> 
>  						John H. Kennedy
> 
We were also using RTAda hosted on a VAX. We decided to use Rational's Apex
and VADScross on Sun/Solaris. At the same time we changed our target system
from 68k board MVME147 to PPC board MVME1604. As an intermediate step we
only switched  the development environment. Then we ungraded to the more
powerful PPC-boards. The results we experienced are as follows:

* If your program doesn't rely on compiler proprietary behavior of data
representation
  (e.g. integer'size = 16, alignment of array elements) plain ada programs
need almost no
  effort for porting.

* ARTX calls (e.g. SC_SPost) are pretty easy to replace by appropriate
calls of the new
  RTS. I would expect that this is true for other RTS, e.g. VxWorks.

* Changing the code for interrupt handling was also not too much work. We
are using
  both procedures and task entries as ISR. (If you use the RTAda hybrid
method for
  interrupt handling porting might be more time consuming)

* Switching the target-HW is what really costs time and money.

                                            Hartmut Schaefer  




      parent reply	other threads:[~1998-01-09  0:00 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
1998-01-08  0:00 Ada compilers and RTOSs jkennedy
1998-01-09  0:00 ` Hartmut H. Schaefer
1998-01-09  0:00 ` Hartmut H. Schaefer [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