netdev - Re: [PATCH] dcache: better name hash function

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <20091027103536.28cf0a0f@nehalam>
Date:	Tue, 27 Oct 2009 10:35:36 -0700
From:	Stephen Hemminger <shemminger@...tta.com>
To:	Eric Dumazet <eric.dumazet@...il.com>,
	David Miller <davem@...emloft.net>
Cc:	netdev@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] dcache: better name hash function

On Tue, 27 Oct 2009 18:19:13 +0100
Eric Dumazet <eric.dumazet@...il.com> wrote:

> Stephen Hemminger a écrit :
> 
> > So Eric's string10 is fastest for special case of fooNNN style names.
> > But probably isn't best for general strings. Orignal function
> > is >20% slower, which is surprising probably because of overhead
> > of 2 shifts and multipy. jenkins and fnv are both 10% slower.
> > 
> 
> 
> jhash() is faster when strings are longer, being able to process 12 bytes per loop.
> 

But jhash is not amenable to usage in namei (with partial_name_hash).

name_hash is rarely done on long strings, the average length of a filename
is fairly short (probably leftover Unix legacy). On my system, average
path component length in /usr is 13 characters; therefore jhash has
no big benefit here.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html