Re: buildd failure for hppa - floating point encoding?

[Helen Faulkner <helen_ml_faulkner@yahoo.co.uk>]

(Please CC replies to me - I'm not on the list)

Matthew Wilcox wrote:
> On Tue, Oct 26, 2004 at 09:50:56AM +0100, Helen Faulkner wrote:
>
> > The program defines the floating point encoding in the included cdf library 
> > like this (from cdf/cdfdist.h, line 437 onwards):
>
>
> Per the responses on the debian-ia64 list, I believe PA-RISC uses the
> IEEE big-endian format.

Ahh, cool.  Do you happen to know what identifier would be defined for a PA-RISC 
machine (I assume this is the same as hppa - I'm sorry to admit I really don't 
know much about that stuff).

Ie, if I were to use:

#if 
defined(MIPSEL)||defined(IBMPC)||defined(alphaosf)||defined(alphavmsI)||defined(posixSHELLalphaI)||defined(something_for_hppa)
#  define FP2cpu
#endif

to get it to know that a hppa machine uses that same encoding.  What would 
something_for_hppa be?

Or would it be the same or better to say, earlier:

#    if defined(i386) || defined(__ia64__) || defined(something_for_hppa)
#      define IBMPC

I guess the latter is better if hppa is the same as IBMPC in ways other than the 
floating point encoding.

> BTW, you'll have fun by the time you get to ARM.  It uses the IEEE format,
> except it has the word order reversed.

Yeah, someone already told me that.  That one may have to go to upstream to deal 
with - I'm already scared of making a change that mangles someone's data somehow.

Thanks very much for helping.

Helen.