1 |
Oh, and euhm.. |
2 |
What about this?: |
3 |
http://gcc.gnu.org/faq.html#optimizing |
4 |
|
5 |
If gcc itself is optimised with for example '-O3 -fomit-frame-pointer', |
6 |
won't that make it faster than gcc 2.95.3? :o) |
7 |
Just taking wild guesses here :o) |
8 |
|
9 |
See ya |
10 |
|
11 |
On Monday 08 April 2002 00:54, Spider wrote: |
12 |
|| Hello, I've just upgraded my -rc6 to -1.0-gcc3 and decided to make an |
13 |
|| (unofficial) benchmark. |
14 |
|| |
15 |
|| I went for galeon, I had originally intended to use mozilla, but the |
16 |
|| time-results borked so I go for galeon instead.. smaller codebase, so |
17 |
|| its not as great difference, but it does have both c and c++ code, so it |
18 |
|| might be a decent choice. |
19 |
|| |
20 |
|| |
21 |
|| gcc 2.95.3 : |
22 |
|| real 3m38.592s |
23 |
|| user 2m46.810s |
24 |
|| sys 0m28.100s |
25 |
|| CFLAGS="-march=i686 -O3 -pipe" |
26 |
|| CXXFLAGS="-march=i686 -O3 -pipe" |
27 |
|| |
28 |
|| |
29 |
|| gcc 3.0.4 : |
30 |
|| real 5m6.465s |
31 |
|| user 3m27.440s |
32 |
|| sys 0m30.140s |
33 |
|| CFLAGS="-march=athlon -O3 -pipe" |
34 |
|| CXXFLAGS="-march=athlon -O3 -pipe" |
35 |
|| |
36 |
|| |
37 |
|| |
38 |
|| |
39 |
|| if you only compare the "user" time it should be enough... as the "sys" |
40 |
|| show, there's a few percentages difference between them, so this is not |
41 |
|| scientific or anything. |
42 |
|| |
43 |
|| Would be interesting to compare the results as well, since those are |
44 |
|| quite likely rather different with the new levels of optimization... |
45 |
|| |
46 |
|| |
47 |
|| //Spider |
48 |
|
49 |
-- |
50 |
Bart Verwilst |
51 |
Gentoo Linux Developer, Desktop Team |
52 |
Gent, Belgium |