Linux kernel bindings - at least sockets?

Linux kernel bindings - at least sockets?

[Le]

I'm looking for and Ada(95) binding for the Linux kernel.  In the short
term I need at least the BSD sockets interfaces - I don't desire Posix
I/F.

Re: Linux kernel bindings - at least sockets?

[Jeffrey Creem]

Le wrote:
> I'm looking for and Ada(95) binding for the Linux kernel.  In the short
> term I need at least the BSD sockets interfaces - I don't desire Posix
> I/F.
> 


Do you really mean that you want to use BSD sockets from within Kernel 
code or do you simply mean that you want to use Sockets from user land 
code on a Linux system.

If the former, you've got bigger issues than simply getting sockets 
interfaces since you need to figure out how to do Ada kernel modules 
(which several people have done).

If the latter, then how about http://www.rfc1149.net/devel/adasockets

Re: Linux kernel bindings - at least sockets?

[Mark Lorenzen]

"Le" <snelson1@llnl.gov> writes:

> I'm looking for and Ada(95) binding for the Linux kernel.  In the short
> term I need at least the BSD sockets interfaces - I don't desire Posix
> I/F.

ISO standard ISO/IEC 14519 defines an Ada API to POSIX. It is a solid
piece of work and a good inspiration to all who wants to write an Ada
API in a world that is traditionally dominated by C.

There exists at least one implementation of ISO/IEC 14519 called
'florist', which is compatible with the GNAT compiler. I have used it
with success on Linux.

- Mark

Re: Linux kernel bindings - at least sockets?

[Le]

Thanks Jeffrey. I had no idea there was an RFC in place for this
interface.

Re: Linux kernel bindings - at least sockets?

[Le]

Thanks Mark. I initially rejected florist because I, probably
erroneously, saw a difference between the wrapping of VxWorks I/F in
our compiler vendors implementation and the Posix I/F in florist. Now
that I have the RFC, thanks to Jeffrey, I will be able to compare the
I/Fs and make a, hopefully, more judicious selection based on our new
target's sockets I/Fs.

Re: Linux kernel bindings - at least sockets?

[Le]

Thanks Mark. I initially rejected florist because I, probably
erroneously, saw a difference between the wrapping of VxWorks I/F in
our compiler vendors implementation and the Posix I/F in florist. Now
that I have the RFC, thanks to Jeffrey, I will be able to compare the
I/Fs and make a, hopefully, more judicious selection based on our new
target's sockets I/Fs.

Re: Linux kernel bindings - at least sockets?

[Simon Wright]

"Le" <snelson1@llnl.gov> writes:

> Thanks Jeffrey. I had no idea there was an RFC in place for this
> interface.

Check out the official RFC at http://www.faqs.org/rfcs/rfc1149.html ..

AdaSockets is one binding, if you are using GNAT check out
GNAT.Sockets which is said to have its problems especially in 3.15 but
which in fact works pretty well for straightforward use.