From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alan Cox Subject: Re: [Patch] Fix oops on rmmod usb-storage Date: Wed, 29 Sep 2004 14:17:58 +0100 Sender: linux-scsi-owner@vger.kernel.org Message-ID: <1096463876.15905.23.camel@localhost.localdomain> References: <415A67B8.2080003@suse.de> <1096466196.2028.8.camel@mulgrave> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: Received: from clock-tower.bc.nu ([81.2.110.250]:6794 "EHLO localhost.localdomain") by vger.kernel.org with ESMTP id S268447AbUI2OUy (ORCPT ); Wed, 29 Sep 2004 10:20:54 -0400 In-Reply-To: <1096466196.2028.8.camel@mulgrave> List-Id: linux-scsi@vger.kernel.org To: James Bottomley Cc: Hannes Reinecke , Linux Kernel Mailing List , Andrew Morton , SCSI Mailing List On Mer, 2004-09-29 at 14:56, James Bottomley wrote: > The key to the solution of this problem is to know what USB is trying to > do with the dead device. SCSI is trying to be polite and explicitly > kill the outstanding commands before it removes the HBA. Presumably USB > is returning something that says this can't be done so the EH gets all > the way up to offlining. Its nothing to do with USB, rmmod with eh running crashes all the other SCSI drivers I've tested too. After the state transition fails you get kobject related errors and a crash. That makes me suspect whatever is ill is in the scsi core. Alan