From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: [PATCH 0/25] Sysfs cleanups & tagged directory support Date: Wed, 08 Aug 2007 17:54:44 +0900 Message-ID: <46B984D4.9020600@gmail.com> References: <11860582832964-git-send-email-htejun@gmail.com> <20070808093815.0935a18e@gondolin.boeblingen.de.ibm.com> <20070808095432.4a520735@gondolin.boeblingen.de.ibm.com> <20070808103759.0850383c@gondolin.boeblingen.de.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20070808103759.0850383c@gondolin.boeblingen.de.ibm.com> Sender: linux-kernel-owner@vger.kernel.org To: Cornelia Huck Cc: "Eric W. Biederman" , Greg KH , linux-kernel@vger.kernel.org, satyam@infradead.org, stern@rowland.harvard.edu, Linux Containers , gregkh@suse.de List-Id: containers.vger.kernel.org Cornelia Huck wrote: > On Wed, 08 Aug 2007 01:57:07 -0600, > ebiederm@xmission.com (Eric W. Biederman) wrote: > >>> Got it: It's patch 6, the readdir simplification. >>> >>> (The udev on that guest is ancient (063)...) >> Ok. That is weird. > > More weirdness. If I activate another dasd from the repair file > system, /dev/dasdb is created... > > Same if I set the card reader online: /dev/vmrdr-0.0.000c is created as > expected. > >> Does it depend on the order in which the dentries are returned from >> readdir? > > I'd think not. > >> Unless I made a really stupid error otherwise the two versions >> of readdir should have the same semantics. > > Yes, your patch looks sane. I have no idea why it breaks stuff... Are you sure it's patch 6? Patch 17 adds a deadlock in rename path. -- tejun