From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Frantisek Rysanek" Subject: Re: Disk IO, "Paralllel sequential" load: read-ahead inefficient? FS tuning? Date: Fri, 10 Apr 2009 08:19:58 +0200 Message-ID: <49DF012E.17197.988E1FE5@localhost> References: <49DCCF55.32475.8FFB6157@localhost> (Frantisek Rysanek's message of "Wed, 08 Apr 2009 16:22:45 +0200") Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7BIT To: linux-fsdevel@vger.kernel.org Return-path: Received: from smtp.seznam.cz ([77.75.72.43]:39449 "EHLO smtp.seznam.cz" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S937695AbZDJGUI (ORCPT ); Fri, 10 Apr 2009 02:20:08 -0400 In-reply-to: <87prfl8zzk.fsf@basil.nowhere.org> Content-description: Mail message body Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On 9 Apr 2009 at 20:05, Andi Kleen wrote: [...] > unsigned long max_sane_readahead(unsigned long nr) [...] > Wu Fengguang (cc'ed) is doing a lot of work on the MADV_* readahead > algorithms. There was a recent new patchkit from him on linux- kernel [...] > For example for faster booting sys_readahead() is widely used now. > Okay, thanks for the pointers :-) Now I know where to get follow-up reading / try some hacking... Frank Rysanek