1 |
So hi from a amd64 newbie. Not so newbie with Gentoo though. :) |
2 |
|
3 |
My system is an Intel quad core core2 with a 2.4 GHz clock speed coupled with |
4 |
a 4GB of memory. No overclocking etc. Want this to be stable. :) |
5 |
|
6 |
I'm just curious what people use as their stable CFLAGS in amd64 Gentoo? |
7 |
(Sorry if this has been up lately, but I just switched to 64bit env so...) |
8 |
|
9 |
|
10 |
Here is mine and some explanation of why (And I use ~arch system with gcc 4.3) |
11 |
|
12 |
The flags are in order they are used in my CFLAGS and CXXFLAGS. |
13 |
|
14 |
Gives stable base |
15 |
-O2 |
16 |
|
17 |
Want to optimize for my system, but don't want "native" |
18 |
-march=core2 |
19 |
|
20 |
If some ebuilds filter march this will still cache optimize etc for my system |
21 |
-mtune=core2 |
22 |
|
23 |
Faster floating point math and better chance of vectorization |
24 |
-mfpmath=sse |
25 |
|
26 |
These because of the march might get filtered |
27 |
-mmmx -msse -msse2 -msse3 -mssse3 |
28 |
|
29 |
For loop vectorization |
30 |
-ftree-vectorize |
31 |
|
32 |
Just to get some Idea how much vectorized loops there will be. |
33 |
By the way I surprised the amount of "LOOP VECTORIZED" notes in the compile |
34 |
output. And only have seen couple of two versions |
35 |
-ftree-vectorizer-verbose=1 |
36 |
|
37 |
Of course I don't want temp files :) |
38 |
-pipe |
39 |
|
40 |
|
41 |
I don't use any loop unrolling etc, because it would only add the code size. |
42 |
I'm not so brave that I would dare to use -Os. |
43 |
|
44 |
So what's your experiences and reasoning behind what you do? |
45 |
Any benchmarks or so? |
46 |
|
47 |
|
48 |
PS. If you see same post without this added postscript. Just ignore it, it's |
49 |
the same post, but I forgot to change my default identity for this ML. |