From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756119AbYIKUIq (ORCPT ); Thu, 11 Sep 2008 16:08:46 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752926AbYIKUIh (ORCPT ); Thu, 11 Sep 2008 16:08:37 -0400 Received: from terminus.zytor.com ([198.137.202.10]:60119 "EHLO terminus.zytor.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751986AbYIKUIh (ORCPT ); Thu, 11 Sep 2008 16:08:37 -0400 Message-ID: <48C979D3.5000308@zytor.com> Date: Thu, 11 Sep 2008 13:04:35 -0700 From: "H. Peter Anvin" User-Agent: Thunderbird 2.0.0.14 (X11/20080501) MIME-Version: 1.0 To: Dean Nelson CC: "Eric W. Biederman" , Alan Mayer , Ingo Molnar , jeremy@goop.org, rusty@rustcorp.com.au, suresh.b.siddha@intel.com, torvalds@linux-foundation.org, linux-kernel@vger.kernel.org, Thomas Gleixner , Yinghai Lu Subject: Re: [RFC 0/4] dynamically allocate arch specific system vectors References: <489C6844.9050902@sgi.com> <20080811165930.GI4524@elte.hu> <48A0737F.9010207@sgi.com> <20080911152304.GA13655@sgi.com> In-Reply-To: <20080911152304.GA13655@sgi.com> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Dean Nelson wrote: > > We (SGI) need somewhere around eight vectors. > > There are two kernel modules, sgi-gru and sgi-xp (in drivers/misc), that > each need two vectors. And there's the broadcast assist unit (BAU) that is > involved in tlb shootdown on uv, which currently uses statically reserved > vector 0xf8 (UV_BAU_MESSAGE -- see uv_bau_init()). I know of a debugger that > also uses 0xf8 because it was previously available until UV_BAU_MESSAGE came > along. The BAU would be happy with a dynamically allocated system vector. > We have a couple of other things in the works that also need vectors. > > All of these eight or so vectors are only meaningful on SGI uv systems. > Are these kernel-internal vectors, or exposed to userspace (i.e. the INT instruction works in userspace)? From what I'm gathering, I think this is the former. There are the occational user who wants fixed user-space addressible vectors, effectively as secondary system call entry vectors (lguest is the only in-kernel user that I know of); those need to be static and non-conflicting, which is kind of difficult. It seems to me that anything else probably should be possible to be dynamic. -hpa