1 |
I think the important question is -- does gcc3 generate better/faster |
2 |
code? If it generates faster code, I'm okay with spending a little more |
3 |
time to create it because over the life of the system, the amount of |
4 |
time I spend compiling is small in comparison. |
5 |
|
6 |
Ric |
7 |
|
8 |
|
9 |
|> -----Original Message----- |
10 |
|> From: gentoo-dev-admin@g.o |
11 |
|> [mailto:gentoo-dev-admin@g.o] On Behalf Of Spider |
12 |
|> Sent: Sunday, April 07, 2002 6:55 PM |
13 |
|> To: gentoo-dev |
14 |
|> Subject: [gentoo-dev] gcc 2.95.3 / 3.0.4 speed comparsion |
15 |
|> |
16 |
|> |
17 |
|> Hello, I've just upgraded my -rc6 to -1.0-gcc3 and decided to make an |
18 |
|> (unofficial) benchmark. |
19 |
|> |
20 |
|> I went for galeon, I had originally intended to use mozilla, |
21 |
|> but the time-results borked so I go for galeon instead.. |
22 |
|> smaller codebase, so its not as great difference, but it |
23 |
|> does have both c and c++ code, so it might be a decent choice. |
24 |
|> |
25 |
|> |
26 |
|> gcc 2.95.3 : |
27 |
|> real 3m38.592s |
28 |
|> user 2m46.810s |
29 |
|> sys 0m28.100s |
30 |
|> CFLAGS="-march=i686 -O3 -pipe" |
31 |
|> CXXFLAGS="-march=i686 -O3 -pipe" |
32 |
|> |
33 |
|> |
34 |
|> gcc 3.0.4 : |
35 |
|> real 5m6.465s |
36 |
|> user 3m27.440s |
37 |
|> sys 0m30.140s |
38 |
|> CFLAGS="-march=athlon -O3 -pipe" |
39 |
|> CXXFLAGS="-march=athlon -O3 -pipe" |
40 |
|> |
41 |
|> |
42 |
|> |
43 |
|> |
44 |
|> if you only compare the "user" time it should be enough... |
45 |
|> as the "sys" show, there's a few percentages difference |
46 |
|> between them, so this is not scientific or anything. |
47 |
|> |
48 |
|> Would be interesting to compare the results as well, since |
49 |
|> those are quite likely rather different with the new levels |
50 |
|> of optimization... |
51 |
|> |
52 |
|> |
53 |
|> //Spider |
54 |
|> |
55 |
|> |
56 |
|> -- |
57 |
|> begin happy99.exe |
58 |
|> This is a .signature virus! Please copy me into your |
59 |
|> .signature! See Microsoft KB Article Q265230 for more |
60 |
|> information. end |
61 |
|> |