[boost] Re: Numeric Conversion Formal Review

From: Guillaume Melquiond (guillaume.melquiond_at_[hidden])
Date: 2003-12-12 12:04:55

Le ven 12/12/2003 à 17:08, Dan W. a écrit :
>  > No need to go as far as using assembly language. There are C functions
>  > that work just fine like nearbyint or rint (the second one is usually
>  > faster). And if you are really looking for assembler, the frndint
>  > instruction is available on x86 for exemple.
> 
> Not familiar with those, but I just came up with something that should
> work...

[... snipped ...]

> -------------------------------------------------------------------
> (sorry if my syntax offends, I'm rusty in assembler these days)
> 
> That should add up to just four integer instructions, with optims' ON.
> Not totally sure that the sign bit would always survive; would need to
> test it.
> 
> Cheers!

Is this some kind of joke? What did you expect when you posted this
mail? Please refrain from going wild on a public-mailing list.

Most processors have an opcode to round a floating-point value to an
integer. For example, as I said, this instruction is called frndint on
x86 (how do you think floor, rint and ceil are coded?). Don't try to
emulate a floating-point unit by using a handful of integer operations.
First it will be completely wrong (it's not as easy as you seem to
believe), and second it will be slow.

Guillaume