Re: Sam Figueroa's thesis.

Re: Sam Figueroa's thesis.

[dewar]

One thing we can do is to make it clear that we are not covering the
entire IEEE standard. We don't care about infinities and NaN's I would
think too much, and we certainly don't care about odd operations like
unordered comparisons, and we certainly don't care about odd rounding modes.

Re: Sam Figueroa's thesis.

[dewar]

<<I found the thesis you reference below online after only a few minutes
of googlin' :-)
>>

Great, glad to hear it is accessible :-)

But as per previous note, I don't think you need to go ploughing through
this right now, for now it is good enough to reference the appropriate
section in C99.

What we need eventually in your report is a list of references on 
floating-point models for each language.

For Ada, we have two sources,

a) Annex G, the official model in the RM, which is very precise, but is
general enough to accomodate typical non-IEEE architectures (for instance
there is permission to use multiply-by-the-reciprocal if the hardware does
this, so that if anyone does ever implement Ada on the Cray or i860, the
definition will be ready for it :-)

b) the IEEE binding that appears in Sam's thesis. Never fully implemented,
but the basic evaluation rules are clear.

Sam Figueroa's thesis.

[Toon Moene]

Hmmm, I must have been very tired, yesterday.

I found the thesis you reference below online after only a few minutes
of googlin' :-)

> But this is meaningless, there *is* no "IEEE-754" model for evaluation of
> floating-point expressions in high level languages. So this model needs
> a lot of filling out. I refer again to Sam Figueroa's PhD thesis which is
> all about such models.

--
Toon Moene, KNMI, PO Box 201, 3730 AE De Bilt, The Netherlands.
Tel. +31302206443, Fax +31302210407,  e-mail moene@knmi.nl
URL: http://www.knmi.nl/hirlam