| 1 |
Hi guys, |
| 2 |
|
| 3 |
Just read an interesting article about Xeon vs. Opteron from anandtech - |
| 4 |
where they really show how much difference compile optimizations (or not) |
| 5 |
does - and how it differs for different programs for different processors. |
| 6 |
|
| 7 |
http://www.anandtech.com/linux/showdoc.aspx?i=2163&p=1 |
| 8 |
|
| 9 |
To me this clearly shows, that if Gentoo wants the best performance - we |
| 10 |
can't use a "one cflags fits them all" approach. I do know that if a |
| 11 |
program breaks, those CFLAGS are pulled out in the individual ebuild, but |
| 12 |
this is not due to poor performance. |
| 13 |
|
| 14 |
IMHO the only way for Gentoo to prove its true potential - is to somehow |
| 15 |
build an array of compile options, with CPU's on X, programs on Y and |
| 16 |
GCC-version on Z. Getting the numbers for each CPU, will ofcourse require |
| 17 |
writing tests, for each program - but IMHO this can be done, if we do it |
| 18 |
one at a time. |
| 19 |
|
| 20 |
I would suggest these tests be included like the gentoo-stats program, as |
| 21 |
something the individual Gentooist can choose to run after each compile - |
| 22 |
which would give him the optimal performance (and recompile X number of |
| 23 |
times to test different flags out) on his CPU/program/GCCversion |
| 24 |
combination, and at the same time, send the result to a Gentoo database. |
| 25 |
|
| 26 |
I know I would definetely have the patience to let it test and test again, |
| 27 |
if it meant more performance for me Smile |
| 28 |
|
| 29 |
The end result should be, that Gentoo automagically selects the optimal |
| 30 |
CFLAGS (in performance and stability - perhaps with some optimizations |
| 31 |
flagged as "unstable" so people can select "optimize for performance" vs. |
| 32 |
"optimize for stability") depending on the X, Y and Z from above. |
| 33 |
|
| 34 |
I would very much like to be one of the guys that gets the ball rolling, |
| 35 |
but as I'm not a Gentoo Dev - We (or just I) need to agree with the Gentoo |
| 36 |
Dev's on how this could best be done. |
| 37 |
|
| 38 |
What do you think? am I crazy? It seems to me that the anandtech tests |
| 39 |
shows that it is more than just a 1% or 2% difference, with the right |
| 40 |
CFLAGS - and that the right CFLAGS for one program, can be the worst for |
| 41 |
another on same CPU/GCC combination. |
| 42 |
|
| 43 |
|
| 44 |
-- |
| 45 |
Regards, |
| 46 |
Klavs Klavsen, GSEC - kl@××××.dk - http://www.vsen.dk |
| 47 |
PGP: 7E063C62/2873 188C 968E 600D D8F8 B8DA 3D3A 0B79 7E06 3C62 |
| 48 |
|
| 49 |
"Those who do not understand Unix are condemned to reinvent it, poorly." |
| 50 |
--Henry Spencer |
| 51 |
|
| 52 |
|
| 53 |
-- |
| 54 |
gentoo-dev@g.o mailing list |
Archives