From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261396AbULXM1N (ORCPT ); Fri, 24 Dec 2004 07:27:13 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261397AbULXM1N (ORCPT ); Fri, 24 Dec 2004 07:27:13 -0500 Received: from pcsmail.patni.com ([203.124.139.197]:43922 "EHLO pcsmail.patni.com") by vger.kernel.org with ESMTP id S261396AbULXM1J convert rfc822-to-8bit (ORCPT ); Fri, 24 Dec 2004 07:27:09 -0500 X-MimeOLE: Produced By Microsoft Exchange V6.5.7226.0 Content-class: urn:content-classes:message MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7BIT Subject: Pseudo Driver hangs :b_elv_sequence ?SUSE 8.0 SP3 Date: Fri, 24 Dec 2004 17:57:07 +0530 Message-ID: <374639AB1012AA4C840022842AA95BC203C70201@ruby.patni.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: What's wrong with Documentation/DocBook/kernel-api.tmpl? (was Re: Something wrong when transform Documentation/DocBook/*.tmpl into pdf) Thread-Index: AcTpr8DiPj5xQBunRsqzO3hhC3fOvAAcxhuw X-Priority: 1 Importance: high From: "Kotian, Deepak" To: Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi, I saw a response from you on b_elv_sequence over here. http://www.uwsg.iu.edu/hypermail/linux/kernel/0301.3/0472.html I have observed a similar problem in 2.4.21-138 kernel which comes along with SUSE 8.0 SP3. This is mainly for SUSE. I presume, SUSE folks are active on this forum and would respond immediately.Please excuse others, sorry. In Our pseudo block driver a Defect occurs only when read is done from urandom device & write to raw device. It basically hangs in kiobuf module. We are using the buffer_head structure. We have found by trial and error and looking the source at of raid and md devices that if b_elv_sequence parameter is set to zero, it is OK and our driver does not have a problem. On looking at the site,we found that you are aware of it. Could you please elaborate. Question: Please let us know your views on the b_elv_sequence and reason to set it to zero. We need to justify whether our fix is correct. Also, this is not there SLES 9.0 with kernel 2.6 as buffer head concept is gone. But is the functionality mapped of this flag mapped somewhere in 2.6 or it is not needed. Anyone here is aware of it. P.S. I have posted on SUSE support, not sure when I get response after internal processing. As it is very urgent for me. I am posting it here as well. Thanks and Regards Deepak