Return-Path: Received: from bombadil.infradead.org ([198.137.202.9]:49094 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753318AbcFQLbL (ORCPT ); Fri, 17 Jun 2016 07:31:11 -0400 Date: Fri, 17 Jun 2016 04:31:10 -0700 From: Christoph Hellwig To: "J. Bruce Fields" Cc: Christoph Hellwig , linux-nfs@vger.kernel.org Subject: Re: odd sillyrename bug in 4.6 / 4.7-rc Message-ID: <20160617113110.GA8864@infradead.org> References: <20160616150146.GA14015@infradead.org> <20160616172952.GA17604@fieldses.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20160616172952.GA17604@fieldses.org> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Thu, Jun 16, 2016 at 01:29:52PM -0400, J. Bruce Fields wrote: > On Thu, Jun 16, 2016 at 08:01:47AM -0700, Christoph Hellwig wrote: > > I've run into an odd bug when testing the clone / reflink functionality, > > and managed to isolate a testcase not relying on clones. The issue is > > that after a lot of direct I/O hitting ENOSPC on a NFS mount I run > > into a silly rename that never gets cleaned up after cleaning up the > > directory. > > I have a redhat bug for what sounds like the same behavior. I spent > some time looking at it but I don't think I got anywere. > > https://bugzilla.redhat.com/show_bug.cgi?id=1132610 > > (But you probably can't see that right now; I'll see if I can open it.) I can't see it indeed.