From: "J. R. Okajima" Subject: Re: NFS, race in writeback? Date: Mon, 24 May 2010 13:56:44 +0900 Message-ID: <28595.1274677004@jrobl> References: <11621.1274548448@jrobl> <1274640976.4860.97.camel@heimdal.trondhjem.org> Cc: linux-nfs@vger.kernel.org, linux-kernel@vger.kernel.org To: Trond Myklebust Return-path: Received: from mtoichi12.ns.itscom.net ([219.110.2.182]:51519 "EHLO mtoichi12.ns.itscom.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1749667Ab0EXE5N (ORCPT ); Mon, 24 May 2010 00:57:13 -0400 In-Reply-To: <1274640976.4860.97.camel-rJ7iovZKK19ZJLDQqaL3InhyD016LWXt@public.gmane.org> Sender: linux-nfs-owner@vger.kernel.org List-ID: Trond Myklebust: > Urgh. Yes, this looks like it is a consequence of commit > ba8b06e67ed7a560b0e7c80091bcadda4f4727a5. We need to revert the part > that calls sync_inode(). > > Does the following fix it for you? Applying the patch against 2.6.34, it passed my local stress test for two hours. The problem was fixed. Thank you. J. R. Okajima