Gentoo Archives: gentoo-dev

From: Spider <spider@××××××××××.net>
To: gentoo-dev@g.o
Subject: Re: [gentoo-dev] gcc 2.95.3 / 3.0.4 speed comparsion
Date: Sun, 07 Apr 2002 18:31:18
Message-Id: 20020408012500.39eeb26f.spider@deathangel.net
In Reply to: [gentoo-dev] gcc 2.95.3 / 3.0.4 speed comparsion by Spider
1 And some more numbers:
2
3 gcc 3.0.4 - gabber
4 real 14m51.179s
5 user 12m3.010s
6 sys 0m34.320s
7
8 gcc 2.95.3 - gabber
9 real 7m53.177s
10 user 6m57.450s
11 sys 0m29.940s
12
13
14
15 begin quote
16 On Mon, 8 Apr 2002 00:54:46 +0200
17 Spider <spider@g.o> wrote:
18
19 > Hello, I've just upgraded my -rc6 to -1.0-gcc3 and decided to make an
20 > (unofficial) benchmark.
21 >
22 > I went for galeon, I had originally intended to use mozilla, but the
23 > time-results borked so I go for galeon instead.. smaller codebase, so
24 > its not as great difference, but it does have both c and c++ code, so
25 > it might be a decent choice.
26 >
27 >
28 > gcc 2.95.3 :
29 > real 3m38.592s
30 > user 2m46.810s
31 > sys 0m28.100s
32 > CFLAGS="-march=i686 -O3 -pipe"
33 > CXXFLAGS="-march=i686 -O3 -pipe"
34 >
35 >
36 > gcc 3.0.4 :
37 > real 5m6.465s
38 > user 3m27.440s
39 > sys 0m30.140s
40 > CFLAGS="-march=athlon -O3 -pipe"
41 > CXXFLAGS="-march=athlon -O3 -pipe"
42 >
43 >
44 >
45 >
46 > if you only compare the "user" time it should be enough... as the
47 > "sys" show, there's a few percentages difference between them, so this
48 > is not scientific or anything.
49 >
50 > Would be interesting to compare the results as well, since those are
51 > quite likely rather different with the new levels of optimization...
52 >
53 >
54 > //Spider
55 >
56 >
57 > --
58 > begin happy99.exe
59 > This is a .signature virus! Please copy me into your .signature!
60 > See Microsoft KB Article Q265230 for more information.
61 > end
62 >
63
64
65 --
66 begin happy99.exe
67 This is a .signature virus! Please copy me into your .signature!
68 See Microsoft KB Article Q265230 for more information.
69 end