2 Aug
2012
2 Aug
'12
5:18 a.m.
On Wed, Aug 1, 2012 at 10:53 PM, Rustom Mody <rustompmody(a)gmail.com> wrote:
Any pointers to why floating point in the kernel is
such a no-no?
I could understand it when floating point happened in co-processors.
FPU register state is expensive to save+restore. no point in doing it all
the time when you hardly ever need to.