Changes to lunar optimize

[Zbigniew Luszpinski]

> > cpu selection in safe mode should be limited to:
> > -i686
> > -native
> > only. The first one is paranoid safe and slow, the second one is
> > really safe and the fastest.
> 
> well, I'd argue that -i586 would be even safer than i686, but in
> reality the only safe optimization is -march=native or nothing else,
> and that's certainly acceptable. So lets (in safe mode) only show:
> 
>     ( ) none
>     (x) native

-march=i586 and below will be less safe. Lunar 32bit is compiled for i686 
target. Going below main arch is not safe however not much. That is why on 
32bit Lunar ( ) none target == -march=pentiumpro == -march=i686. You can 
test this using empty export TESTFLAGS="" parameter in script presented in 
PoC.

man gcc says:
pentiumpro
Intel PentiumPro CPU.

i686
Same as "generic", but when used as "march" option, PentiumPro instruction 
set will be used, so the code will run on all i686 family chips.

> > On Lunar general ML I presented a proof that shows native target is
> > better than manually selecting CPU. If newbie will choose bad cpu
> > broken system will be compiled.
> 
> a script will always know best :)
> 
> Also, in your POC your compiler isn't enabling -fpmath=... I wonder
> why, perhaps that's different per CPU whether it's worth it or not.

mfpmath is enabled by default: man gcc - for 32bit =387 for 64bit =sse.
This test script shows only settings set via gcc parameter manually or 
march= option which expands itself to other flags like msse m3dnow or mmmx  
It does not show what is set by default.

have a nice day,
Zbigniew Luszpinski