| 1 |
On Sunday 04 September 2005 05:27, waltdnes@××××××××.org wrote: |
| 2 |
|
| 3 |
> |
| 4 |
> That list shows pentium4, mmx, sse, and sse2. Also, if you have *ANY* |
| 5 |
> version of sse available, you can improve performance by running floating |
| 6 |
> point math via sse, rather than 387 instructions. I recommend... |
| 7 |
> |
| 8 |
> CFLAGS="-O2 -pipe -fomit-frame-pointer -march=pentium4 -mmmx -msse -msse2 |
| 9 |
> -mfpmath=sse" |
| 10 |
> |
| 11 |
> |
| 12 |
|
| 13 |
emm. I would not do this. |
| 14 |
|
| 15 |
|
| 16 |
-mfpmath=sse seems to be slower than -fpmath=387 |
| 17 |
|
| 18 |
http://www.anandtech.com/mac/showdoc.aspx?i=2436&p=5 |
| 19 |
|
| 20 |
has the numbers/made the experience. |
| 21 |
|
| 22 |
It seems, that gcc is not he best optimizer in the world ;) |
| 23 |
-- |
| 24 |
gentoo-user@g.o mailing list |
Archives