From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dominique Martinet Subject: Re: Race condition introduced in 4bf46a27 VFS: Impose ordering on accesses of d_inode and d_flags Date: Thu, 23 Jul 2015 11:43:50 +0200 Message-ID: <20150723094350.GA3227@u-michard> References: <20150722154519.GA20808@u-michard> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Cc: David Howells , Al Viro , Aurelien CEDEYN To: Return-path: Received: from oxalide-out.extra.cea.fr ([132.168.224.8]:53923 "EHLO oxalide-out.extra.cea.fr" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752694AbbGWKms (ORCPT ); Thu, 23 Jul 2015 06:42:48 -0400 Received: from pisaure.intra.cea.fr (pisaure.intra.cea.fr [132.166.88.21]) by oxalide.extra.cea.fr (8.14.2/8.14.2/CEAnet-Internet-out-2.3) with ESMTP id t6N9i6dg012895 for ; Thu, 23 Jul 2015 11:44:06 +0200 Received: from pisaure.intra.cea.fr (localhost [127.0.0.1]) by localhost (Postfix) with SMTP id 9C01F204A34 for ; Thu, 23 Jul 2015 11:47:42 +0200 (CEST) Received: from muguet1.intra.cea.fr (muguet1.intra.cea.fr [132.166.192.6]) by pisaure.intra.cea.fr (Postfix) with ESMTP id 896A8204A0A for ; Thu, 23 Jul 2015 11:47:42 +0200 (CEST) Received: from zia.cdc.esteban.ctsi (out.dam.intra.cea.fr [132.165.76.10]) by muguet1.intra.cea.fr (8.13.8/8.13.8/CEAnet-Intranet-out-1.2) with SMTP id t6N9i5x7012675 for ; Thu, 23 Jul 2015 11:44:05 +0200 Content-Disposition: inline In-Reply-To: <20150722154519.GA20808@u-michard> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: Dominique Martinet wrote on Wed, Jul 22, 2015 at 05:45:19PM +0200: > Test run: > make -j17 on a kernel over 9P/RDMA (nfs-ganesha serving a tmpfs > directory) > > This patch has nothing 9p specific so I'll try to reproduce with virtio > or another filesystem tomorrow. I think 9P hits this more easily because > the lack of cache makes it create and delete inodes alot... building a kernel on a 9p/virtio mount (tried tmpfs underneath first, but also fails with xfs on a "slow" disk) reproduced this as well, should be easier to play with than getting RDMA. It's actually failing alot faster, so will try to get simpler reproducers with that again. no specific mount option. -- Dominique