[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Linux 3.10-2-m68k FPU support



Geert Uytterhoeven dixit:

>Hmm, yes, it does look scary... But as usual, there's even less time to
>update the docs than to update the code ;-)

Thanks for volunteering to do that, though ;-)


Geert Uytterhoeven dixit:

>FPU emulation should be fine, AFAIK (never used it myself, though).

Michael Schmitz dixit:

> The docs are quite clearly out of date. From memory, there may be the odd FPU
[…]
> Seeing as this code is crucial, I'd opt for putting it back in, and finding

These two statements are “it should be fine but I don’t really know”
and “this is important so put it back in, but I don’t remember details”.

Together with what *is* currently in the documentation, this is
*not* reassuring, and I *know* that FPU emulation code all across
OSes (kernels and not) is a sore topic (even being kicked from
some of them).

My request for a clear statement still stands.

On the other hand, I cannot help but wonder how well it compares
to whatever ARAnyM passes as FPU. I know ARAnyM doesn’t handle the
80/64bit precision switch at all (which is why we opted to not use
that for Python), but I smell some sort of knob here… if the knob
can be used to tweak between speed and correctness, I’d opt for
correctness. (OK, BSD user speaking.) It’d probably be faster at
runtime to fix ARAnyM instead but that’s a different upstream.

bye,
//mirabilos
-- 
<diogenese> Beware of ritual lest you forget the meaning behind it.
<igli> yeah but it means if you really care about something, don't
    ritualise it, or you will lose it. don't fetishise it, don't
    obsess. or you'll forget why you love it in the first place.


Reply to: