Gareth Hughes wrote:
>
> After discussions with Brian and Josh, I'm adding an fnclex instruction
> to the END_FAST_MATH macro to clear any FP exceptions that may have
> occurred in the FAST_MATH block. It seems pointless to have a
> GCC-specific END_FAST_MATH, when I can convert the entire block to
> inline assembly and get rid of the _FPU_SET* checks etc.
>
> Do people have strong objections to this? I'm essentially adding GCC as
> a requirement to enable the fast math stuff, but with our change in
> policy regarding FP exceptions this seems reasonable -- and we can
> always add non-GCC code for other Linux-x86 targets as required.
I've changed the linux/x86 requirement to GCC/x86, which will allow
other OSes running on x86 that compile Mesa with GCC to benefit from the
FAST_MATH code (perhaps the *BSDs?). Nice and clean, in any case.
-- Gareth
|
