From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262737AbTDRBuN (ORCPT ); Thu, 17 Apr 2003 21:50:13 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262749AbTDRBuN (ORCPT ); Thu, 17 Apr 2003 21:50:13 -0400 Received: from palrel11.hp.com ([156.153.255.246]:5560 "EHLO palrel11.hp.com") by vger.kernel.org with ESMTP id S262737AbTDRBuM (ORCPT ); Thu, 17 Apr 2003 21:50:12 -0400 Date: Thu, 17 Apr 2003 19:02:07 -0700 From: David Mosberger Message-Id: <200304180202.h3I227mw032608@napali.hpl.hp.com> To: linux-kernel@vger.kernel.org Cc: linux-ia64@linuxia64.org Subject: USB deadlock in v2.5.67 X-URL: http://www.hpl.hp.com/personal/David_Mosberger/ Reply-To: davidm@hpl.hp.com Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org So my 2.5.67 kernel was humming along nicely when I got the idea to unplug a USB keyboard. Instant deadlock. The backtrace shows this: Call Trace: [] hcd_free_dev+0x140/0x240 [] usb_release_dev+0x100/0x140 [] device_release+0x80/0xa0 [] kobject_cleanup+0x100/0x120 [] urb_unlink+0x110/0x1a0 [] usb_hcd_giveback_urb+0x30/0x1a0 [] dl_done_list+0x230/0x2a0 [] ohci_irq+0x290/0x340 [] usb_hcd_irq+0x80/0x100 [] handle_IRQ_event+0xa0/0x120 [] do_IRQ+0x360/0x460 [] ia64_handle_irq+0x70/0x140 [] ia64_leave_kernel+0x0/0x240 translates into line 1249 in hcd.c, where it does: spin_lock_irqsave (&hcd_data_lock, flags); The deadlock is pretty obvious: the same lock has already been acquired urb_unlink(), 4 levels up in the call-chain. Anybody have a fix for this? --david