From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S268038AbUHFAmq (ORCPT ); Thu, 5 Aug 2004 20:42:46 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S268029AbUHFAmq (ORCPT ); Thu, 5 Aug 2004 20:42:46 -0400 Received: from cliff.cse.wustl.edu ([128.252.166.5]:61950 "EHLO cliff.cse.wustl.edu") by vger.kernel.org with ESMTP id S268034AbUHFAmD (ORCPT ); Thu, 5 Aug 2004 20:42:03 -0400 From: Berkley Shands Date: Thu, 5 Aug 2004 19:41:50 -0500 (CDT) Message-Id: <200408060041.i760fn60000117668@mudpuddle.cs.wustl.edu> To: marcelo.tosatti@cyclades.com, wli@holomorphy.com Subject: Re: Severe I/O performance regression 2.6.6 to 2.6.7 or 2.6.8-rc3 Cc: berkley@cse.wustl.edu, linux-kernel@vger.kernel.org Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I took the 2.6.6-bk7 image, and replaced mm/readahead.c and mm/vmscan.c from the 2.6.6-bk6 image (just those two files), and the read ahead error has vanished. However, the kernel panic'ed when reading a 16gb file. It may be related to an ongoing issue with pci-x and scsi error recovery on the x86_64, so until I get into the office, I will not be able to see what's on the console. So clearly the code in readahead.c and vmscan.c in -bk7 is the source of one regression. I'll keep looking at the second bug in the morning. Thanks to all for the pointers on where to look. berkley