From: Andi Kleen <ak@suse.de>
To: "David S. Miller" <davem@redhat.com>
Cc: ak@suse.de, anton@samba.org, mingo@elte.hu, linux-kernel@vger.kernel.org
Subject: Re: speed difference between using hard-linked and modular drives?
Date: Fri, 9 Nov 2001 07:39:46 +0100 [thread overview]
Message-ID: <20011109073946.A19373@wotan.suse.de> (raw)
In-Reply-To: <p731yj8kgvw.fsf@amdsim2.suse.de> <20011109110532.B6822@krispykreme> <20011109064540.A13498@wotan.suse.de> <20011108.220444.95062095.davem@redhat.com>
In-Reply-To: <20011108.220444.95062095.davem@redhat.com>; from davem@redhat.com on Thu, Nov 08, 2001 at 10:04:44PM -0800
On Thu, Nov 08, 2001 at 10:04:44PM -0800, David S. Miller wrote:
> From: Andi Kleen <ak@suse.de>
> Date: Fri, 9 Nov 2001 06:45:40 +0100
>
> Sounds like you need a better hash function instead.
>
> Andi, please think about the problem before jumping to conclusions.
> N_PAGES / N_CHAINS > 1 in his situation. A better hash function
> cannot help.
I'm assuming that walking on average 5-10 pages on a lookup is not too big a
deal, especially when you use prefetch for the list walk. It is a tradeoff
between a big hash table thrashing your cache and a smaller hash table that
can be cached but has on average >1 entries/buckets. At some point the the
smaller hash table wins, assuming the hash function is evenly distributed.
It would only get bad if the average chain length would become much bigger.
Before jumping to real conclusions it would be interesting to gather
some statistics on Anton's machine, but I suspect he just has an very
unevenly populated table.
-Andi
next prev parent reply other threads:[~2001-11-09 6:40 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <Pine.LNX.4.33.0111081802380.15975-100000@localhost.localdomain.suse.lists.linux.kernel>
[not found] ` <Pine.LNX.4.33.0111081836080.15975-100000@localhost.localdomain.suse.lists.linux.kernel>
2001-11-08 23:00 ` Andi Kleen
2001-11-09 0:05 ` Anton Blanchard
2001-11-09 5:45 ` Andi Kleen
2001-11-09 6:04 ` David S. Miller
2001-11-09 6:39 ` Andi Kleen [this message]
2001-11-09 6:54 ` Andrew Morton
2001-11-09 7:17 ` David S. Miller
2001-11-09 7:16 ` Andrew Morton
2001-11-09 8:21 ` Ingo Molnar
2001-11-09 7:35 ` Andrew Morton
2001-11-09 7:44 ` David S. Miller
2001-11-09 7:24 ` David S. Miller
2001-11-10 4:56 ` Anton Blanchard
2001-11-10 5:09 ` Andi Kleen
2001-11-10 13:44 ` David S. Miller
2001-11-10 13:52 ` David S. Miller
2001-11-10 14:29 ` Numbers: ext2/ext3/reiser Performance (ext3 is slow) Oktay Akbal
2001-11-10 14:47 ` arjan
2001-11-10 17:41 ` Oktay Akbal
2001-11-10 17:56 ` Arjan van de Ven
2001-11-15 17:24 ` Stephen C. Tweedie
2001-11-12 16:59 ` [patch] arbitrary size memory allocator, memarea-2.4.15-D6 Ingo Molnar
2001-11-12 18:19 ` Jeff Garzik
2001-11-12 23:26 ` Ingo Molnar
2001-11-13 15:59 ` Riley Williams
2001-11-14 20:49 ` Tom Gall
2001-11-15 1:11 ` Anton Blanchard
2001-11-17 18:00 ` Eric W. Biederman
2001-11-10 13:29 ` speed difference between using hard-linked and modular drives? David S. Miller
2001-11-09 7:14 ` David S. Miller
2001-11-09 7:16 ` David S. Miller
2001-11-09 12:54 ` David S. Miller
2001-11-09 13:15 ` Philip Dodd
2001-11-09 13:17 ` Andi Kleen
2001-11-09 13:25 ` David S. Miller
2001-11-09 13:39 ` Andi Kleen
2001-11-09 13:41 ` David S. Miller
2001-11-09 13:26 ` David S. Miller
2001-11-09 20:45 ` Mike Fedyk
2001-11-09 12:59 ` Alan Cox
2001-11-10 5:20 ` Anton Blanchard
2001-11-09 3:12 ` Rusty Russell
2001-11-09 5:59 ` Andi Kleen
2001-11-09 11:16 ` Helge Hafting
2001-11-12 23:23 ` David S. Miller
2001-11-12 23:14 ` Rusty Russell
2001-11-13 1:30 ` Mike Fedyk
2001-11-13 1:15 ` David Lang
2001-11-12 9:59 ` Rusty Russell
2001-11-08 16:01 Roy Sigurd Karlsbakk
2001-11-08 17:02 ` Ingo Molnar
2001-11-08 17:37 ` Ingo Molnar
2001-11-08 23:59 ` Anton Blanchard
2001-11-09 5:11 ` Keith Owens
2001-11-10 3:35 ` Anton Blanchard
2001-11-10 7:26 ` Keith Owens
2001-11-08 17:53 ` Robert Love
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20011109073946.A19373@wotan.suse.de \
--to=ak@suse.de \
--cc=anton@samba.org \
--cc=davem@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®