From mboxrd@z Thu Jan 1 00:00:00 1970 From: James Bottomley Subject: Re: [linux-usb-devel] Fw: [Bug 3466] New: Bug while connecting USB-HDD (fwd) Date: 26 Sep 2004 18:05:40 -0400 Sender: linux-scsi-owner@vger.kernel.org Message-ID: <1096236347.10924.97.camel@mulgrave> References: Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: Received: from stat16.steeleye.com ([209.192.50.48]:5311 "EHLO hancock.sc.steeleye.com") by vger.kernel.org with ESMTP id S264299AbUIZWFu (ORCPT ); Sun, 26 Sep 2004 18:05:50 -0400 In-Reply-To: List-Id: linux-scsi@vger.kernel.org To: Alan Stern Cc: SCSI development list , Andrew Morton , USB development list On Sun, 2004-09-26 at 17:06, Alan Stern wrote: > Andrew sent this to linus-usb-devel, but it should have gone to > linux-scsi. There's not enough information to say why it happened. However, all the SCSI code checks out (it's dated ... open coded reference counting instead of kref, but it looks sound). The scenario described could be seen if there's a problem in the host reference counting. In that case, there should have been a slab error earlier on in the logs at the point the error occurred saying something like "slab error in kmem_cache_destory(): can't free all objects" It's possible this could be caused by a refcounting race on the commands. James