1 |
> We may be getting a P4 hyperthreaded system to play with, but under |
2 |
> different rules, where I get to do a bit of tuning first. I have |
3 |
> already built the core system on another machine using gcc-3.2.3, |
4 |
> "-march=pentium4 -O3 -pipe -fomit-frame-pointer" I note that the |
5 |
> pentium4 warning still appears in make.conf, though I believe it no |
6 |
> longer applies to this gcc. |
7 |
> |
8 |
> A while ago I emailed this list and asked for information on tests and |
9 |
> settings for HT P4's, without a reply. So again, has anyone done any |
10 |
> tests on a HT P4 and is willing to support the flags they chose as being |
11 |
> "the best"? In particular, does -ffast-math give a measurable gain? |
12 |
> Most of my machines have been built as scientific stations, so accuracy |
13 |
> is more important than ultimate speed, so this is one I have never |
14 |
> tested. I am not interested in the -O9 -max-everything kiddies who have |
15 |
> been so vocal, but reasoned choices. |
16 |
> |
17 |
> If you want to flame, go ahead - but support your statements! |
18 |
|
19 |
Well all packages that cannot handle -march=pentium4 should filter it out. |
20 |
This filtering is imprefect so adding -mcpu=pentium4 should ensure correct |
21 |
scheduling of all packages. One thing, do not use -ffast-math, as can be seen |
22 |
in the manual this can actually produce mathematic errors. Normally one does |
23 |
not want this in most packages, so leave it alone. It could be used in things |
24 |
like games, wher the precision does not matter that much, but it is not safe |
25 |
as a default. |
26 |
|
27 |
Paul |
28 |
|
29 |
-- |
30 |
Paul de Vrieze |
31 |
Gentoo Developer |
32 |
Mail: pauldv@g.o |
33 |
Homepage: http://www.devrieze.net |