Return-Path: linux-nfs-owner@vger.kernel.org Received: from casper.infradead.org ([85.118.1.10]:51171 "EHLO casper.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750854AbaHUQHS (ORCPT ); Thu, 21 Aug 2014 12:07:18 -0400 Received: from ip-64-134-168-64.public.wayport.net ([64.134.168.64] helo=localhost) by casper.infradead.org with esmtpsa (Exim 4.80.1 #2 (Red Hat Linux)) id 1XKUtQ-00008k-Ea for linux-nfs@vger.kernel.org; Thu, 21 Aug 2014 16:07:16 +0000 From: Christoph Hellwig To: linux-nfs@vger.kernel.org Subject: Date: Thu, 21 Aug 2014 11:09:16 -0500 Message-Id: <1408637375-11343-1-git-send-email-hch@lst.de> In-Reply-To: References: Sender: linux-nfs-owner@vger.kernel.org List-ID: This series fixes a wide range of issues with the pnfs block layout driver. Before this we were seeing frequent silent data corruption, softlockups and kernel crashes when running both user applications and test cases like xfstests. After this rewrite of the I/O path we've sorted out all issues under normal operations, although error handling in the block layout driver and its interaction with the core nfs and pnfs code still needs further work. It was pointed out that the requirement for blocksizes smaller or equal to the page size conflicts with existing servers. From the state of the code I'm very sure it was never used for real against those servers, but as soon as someone signs up to testing and fix support for these servers I'm happy to land a helping hand, although rock solid support probably is imposible. This work was sponsored by NetApp, Inc. Changes since V1: - added two more layout stateid handling fixes - change the layoutget path so that the layout driver is responsible for freeing the spliced in payload.