1 |
-----BEGIN PGP SIGNED MESSAGE----- |
2 |
Hash: SHA1 |
3 |
|
4 |
Anthony E. Caudel wrote: |
5 |
> I have an AMD 64x2 with MAKEOPTS set to -j3. |
6 |
> |
7 |
> I have noticed that some compiles (glibc for example) will use the |
8 |
> system to the fullest. top reports %CPU near 0%. Other compiles, such |
9 |
> as openoffice seem to use only half as much cpu time. For this top |
10 |
> reported consistently near 50%. |
11 |
> |
12 |
> So I'm wondering if the openoffice compile used only one of the amd |
13 |
> cores. Do some of the emerges ignore the MAKEOPTS setting? Or is this |
14 |
> an upstream issue with the compile not using threads? |
15 |
> |
16 |
> Tony |
17 |
|
18 |
Tony, |
19 |
Some ebuilds filter out MAKEOPTS="-jX" because the makefile(?) is not |
20 |
written correctly to compile with more than one concurrent compile |
21 |
process. According to the OOo ebuild, -jX is not filtered, so that is |
22 |
not the case. Also, you may have checked the CPU usage when it was |
23 |
extracting the tarball, obviously the choke point here would be the HD, |
24 |
not the CPU. |
25 |
|
26 |
Just some extra info for you, I am not the ebuild expert. HTH |
27 |
|
28 |
- -- |
29 |
Jeremy Olexa |
30 |
(olexa@××××××.edu) |
31 |
Office: EE/CS 1-201 |
32 |
CS/IT Systems Staff |
33 |
University of Minnesota |
34 |
|
35 |
-----BEGIN PGP SIGNATURE----- |
36 |
Version: GnuPG v1.4.4 (GNU/Linux) |
37 |
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org |
38 |
|
39 |
iD8DBQFEuwRJFN7pD9kMi/URAsCSAJwMXoPA1fqg5CHkJxbId0OYsCXEcACeKXGc |
40 |
r9FcluQ/XA0fRm8nBzTzmK4= |
41 |
=goWq |
42 |
-----END PGP SIGNATURE----- |
43 |
-- |
44 |
gentoo-user@g.o mailing list |