[Numpy-discussion] Solaris Sparc build broken

[Charles R Harris]

On Sun, Nov 8, 2009 at 10:11 PM, Charles R Harris <charlesr.harris at gmail.com
> wrote:

>
>
> On Sun, Nov 8, 2009 at 9:43 PM, David Cournapeau <cournape at gmail.com>wrote:
>
>> On Mon, Nov 9, 2009 at 12:49 PM, Charles R Harris
>> <charlesr.harris at gmail.com> wrote:
>>
>> >>
>> >
>> > Umm, PPC doesn't have ieee quad, it is an amalgam of two doubles.
>>
>> Good catch ! I added a new type define for this case, with the
>> corresponding union.
>>
>> I have quickly tested it under rosetta, and it seems that
>> npy_nextafterl gives a reasonable answer for the few values I have
>> tried (i.e. the same as for a double).
>>
>>
> I don't remember how PPC does it's arithmetic with the two doubles, but one
> holds small values that are added to the other double, so I don't think it
> can be just like a double in some circumstances. What's rosetta?
>
>
OTOH, it does look like extended precision on the PPC mac might be doubles:

     The libm functions declared in math.h provide mathematical
library func-tions functions
     tions in single-, double-, and extended-precision IEEE-754 floating-point
     formats on Intel macs, and in single- and double-precision IEEE-754
     floating-point formats on PowerPC macs.

Is that what you see?

 Chuck
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/numpy-discussion/attachments/20091108/4a2e80e2/attachment.html>