AW: [avr-gcc-list] Precision of 64-bit using 32-bit numbers

[Haase Bjoern (PT-BEU/EMT) *]

-----Ursprüngliche Nachricht-----
Von: address@hidden [mailto:address@hidden Im Auftrag von Ben L. Titzer
Gesendet: Dienstag, 18. Januar 2005 19:01
Cc: address@hidden
Betreff: Re: [avr-gcc-list] Precision of 64-bit using 32-bit numbers
>> Currently there is no abililty to do 64-bit (double) floating point 
>> math in avr-libc.
>> Somebody would have to volunteer to add this to avr-libc; 
>> contributions are welcome.
>>
>
>If you are looking for 64-bit integer support, you may try using "long 
>long" as the type.
>
>-B
>

But expect it to be extremely inefficient in the present form. If you'd supply 
subroutines for dividing and multiplying long longs, that could be rapidly 
inserted into gcc. I'm just working on the long long issue.

Yours,

Björn