From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Mon, 17 Jun 2002 11:54:17 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Mon, 17 Jun 2002 11:54:16 -0400 Received: from parcelfarce.linux.theplanet.co.uk ([195.92.249.252]:16901 "EHLO www.linux.org.uk") by vger.kernel.org with ESMTP id ; Mon, 17 Jun 2002 11:54:15 -0400 Date: Mon, 17 Jun 2002 16:54:17 +0100 From: Matthew Wilcox To: linux-kernel@vger.kernel.org Subject: Using tq_immediate Message-ID: <20020617165417.Y9435@parcelfarce.linux.theplanet.co.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5.1i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org If you want to use tq_immediate, here's the current way to do it: queue_task(&bcs->tqueue, &tq_immediate); mark_bh(IMMEDIATE_BH); There's about 80 drivers in the tree which do this, FYI. This sucks, because it sets up a linkage between task queues and bottom halves. If you, like I, wish for bottom halves to go away, we're going to have to change this. Some people want to eradicate task queues too -- and while I think this is a fine idea, I can't see it happening in the next week or so. My proposal is: queue_task(&bcs->tqueue, &tq_immediate); wake_immediate_queue(); While this is a little magic (there's no obvious connection between the two), and it's special-purpose (no equivalent for other task queues), I think it's an acceptable compromise that doesn't require taking out all 80 drivers and shooting them. -- Revolutions do not require corporate support.