perfect MAX_ORDER?

From: imel96@trustix.co.id
Date: Thu Apr 12 2001 - 09:46:50 EST


hi,

the default MAX_ORDER is 10. as i don't know anything about
page usage, i did some tests to see how it affects performance
(with the infamous kernel compile time).
here it is (all +6m):

        1st test 2nd test mean
2 40.153
3 38.543
4 38.065 38.615 38.350
5 36.778 38.696 37.737
6 37.902 37.800 37.851
7 36.990 36.650 36.820
8 37.157 36.379 36.768
9 37.215
10 37.951
11 36.889
12 36.773
13 36.765
14 36.533
15 37.683

so i conclude that the test is inconclusive. only the first
three (MAX_ORDER == [234]) have noticeable difference, but
those must be affected by the shrinked dentry & page cache
hash table size, right?

the machine has 128mb ram, so the maximum MAX_ORDER that really
works is 14 (largest chunk 32mb).

if i use MAX_ORDER = 6 i save one page. so what should i use?
6, 14, or the default (10)?

would somebody care how the chunks used?

                imel

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Sun Apr 15 2001 - 21:00:18 EST