1 |
On Wed, Oct 12, 2005 at 12:01:12AM +0200, Bastian Balthazar Bux wrote: |
2 |
> Sorry, but here the results are not those expected: |
3 |
.51.22 vs .53_rc5... try with a vanilla .53_rc5 please |
4 |
|
5 |
|
6 |
> ==== time emerge --metadata; 1st run; 2.0.51.22-r3 |
7 |
> real 2m24.419s |
8 |
> user 0m12.329s |
9 |
> sys 0m3.644s |
10 |
> |
11 |
> ==== time emerge --metadata; 2nd run; 2.0.51.22-r3 |
12 |
> real 1m17.700s |
13 |
> user 0m12.257s |
14 |
> sys 0m2.976s |
15 |
> |
16 |
> ==== time emerge --metadata; 1st run; 2.0.53_rc5 patched |
17 |
> real 3m14.073s |
18 |
> user 0m12.917s |
19 |
> sys 0m9.433s |
20 |
> |
21 |
> ==== time emerge --metadata; 2nd run; 2.0.53_rc5 patched |
22 |
> real 3m42.874s |
23 |
> user 0m12.869s |
24 |
> sys 0m9.333s |
25 |
|
26 |
Wasn't expecting a massive improvement, although wasn't sure as hell |
27 |
wasn't expecting a 3x increase in sys. :) |
28 |
|
29 |
Should've seen a large tweak for the first .53_rc5 run also, since it |
30 |
(essentially) would be a forced rewrite of the cache due to INHERITED |
31 |
vs _eclasses_ key changes (moving eclass_cache into the backend). |
32 |
|
33 |
Not running anything additional via /etc/portage/modules I'd bet, but |
34 |
asking to verify also... |
35 |
|
36 |
Meanwhile, thanks for testing; contrary to other results, but _any_ |
37 |
regression I'm after. |
38 |
~harring |