RPC/XDR experience?

[linas at austin.ibm.com]

On Fri, Jun 06, 2003 at 05:17:28PM -0500, Olof Johansson wrote:
> linas at austin.ibm.com wrote:
> > Anyone out there with experience with RPC on 64 bit machines?
> >
> > I've got a 'hello-world' RPC program that falls apart when
> > compiled & linked to 64-bit glibc on PPC64.  It doesn't like
> > long ints.  I'm getting the feeling that I may be opening a
> > pandora's box; I'm hoping that someone will reply and state
> > that no, 64-bit RPC's work great on alphas and sparcs and mips,
> > and not to worry ...
>
> Notice however, that the implementation of xdr_long WILL fail for cases where it's found that the
> 64-bit long contains larger (or smaller, for negative) values than can be stored in the 32-bit
> over-the-wire long. This might give surprising results when you're using xdr_long to encode an
> unsigned integer.


This is exactly what is happening.


>    if (xdrs->x_op == XDR_ENCODE
>        && (sizeof (int32_t) == sizeof (long)
>            || (int32_t) *lp == *lp))             //   <--  HERE is the check

OK, get ready:

I presume you pulled this from glibc sunrpc subdir?  Or the kernel
rpc subdir?  (The linux kernel rpc's can/should be different than
the glibc ones).

Is this a 'well known limitation'?   (i.e. that xdr_long will only
send 32 bits), or is this a 'buggy implementation'?

If its 'well known', where is it documented? 'man xdr' gives no hint,
and implies the opposite is true  ...

Does RPC define a 64-bit encoding (that can be received by 32-bit
servers)?

If so, shouldn't the xdr_long be modified to use the 64-bit marshalling?

I'm thinking to myself that there are valid reasons for applications
on 32-bit CPU's to use 64-bit ints.  I know gcc handles 64-bit 'long-long
ints' on intel machines just fine, and has done so for at least 5 years.
(if not 10).  By extension, one must conclude that there are valid
reasons for RPC's to support 64-bit ints, even on 32-bit arch's.
If you buy the argument in the last 3 sentances, then I would think
that xdr_long should be generating a 64-bit wire protocol on ppc64.

There. Pandora's box is now open ...

--linas


** Sent via the linuxppc64-dev mail list. See http://lists.linuxppc.org/