lkml.org 
[lkml]   [2003]   [Mar]   [11]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: Dcache hash distrubition patches
>> Conclusion: the hash distribution (for this simple test) looks fine
>> to me.
>
> Yes, because of the overkill size of the hash table. With a 100K + entry
> table you can make near every hash function look good ;)
>
> Try to reduce it to a smaller number of buckets and see what happens.

OK, after I've stopped being an idiot, and misreading the code, I have
some numbers. They still look pretty good to me. I shrunk us from
1,048,576 buckets to 65536, and loaded 1,150,000 entries in there.



5 3
9 4
44 5
113 6
243 7
519 8
1059 9
1613 10
2458 11
3506 12
4515 13
5349 14
6071 15
6328 16
6369 17
5862 18
5228 19
4305 20
3546 21
2613 22
1981 23
1382 24
928 25
602 26
368 27
230 28
115 29
75 30
45 31
16 32
14 33
3 34
4 35
2 36

It's not perfect, but not bad either. Some mathematician can go calculate
just how imperfect it is over random distribution, but it looks OK to me ;-)

M

-
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/

\
 
 \ /
  Last update: 2005-03-22 13:33    [W:0.037 / U:0.580 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site