From owner-freebsd-arch@FreeBSD.ORG Tue May 27 08:54:57 2008 Return-Path: Delivered-To: arch@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 4FA631065672; Tue, 27 May 2008 08:54:57 +0000 (UTC) (envelope-from rwatson@FreeBSD.org) Received: from cyrus.watson.org (cyrus.watson.org [209.31.154.42]) by mx1.freebsd.org (Postfix) with ESMTP id 1E56C8FC18; Tue, 27 May 2008 08:54:57 +0000 (UTC) (envelope-from rwatson@FreeBSD.org) Received: from fledge.watson.org (fledge.watson.org [209.31.154.41]) by cyrus.watson.org (Postfix) with ESMTP id A01BB46B2D; Tue, 27 May 2008 04:54:56 -0400 (EDT) Date: Tue, 27 May 2008 09:54:56 +0100 (BST) From: Robert Watson X-X-Sender: robert@fledge.watson.org To: Poul-Henning Kamp In-Reply-To: <45633.1211870269@critter.freebsd.dk> Message-ID: <20080527095317.R35026@fledge.watson.org> References: <45633.1211870269@critter.freebsd.dk> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Cc: Ed Schouten , current@FreeBSD.org, arch@FreeBSD.org, Ade Lovett , net@FreeBSD.org, "Bruce M. Simpson" Subject: Re: HEAD UP: non-MPSAFE network drivers to be disabled (was: 8.0 network stack MPsafety goals (fwd)) X-BeenThere: freebsd-arch@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussion related to FreeBSD architecture List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 27 May 2008 08:54:57 -0000 On Tue, 27 May 2008, Poul-Henning Kamp wrote: > In message <20080527062736.GF64397@hoeg.nl>, Ed Schouten writes: > >>>> The new TTY layer should still support line disciplines, [...] >>> >>> Actually... I think it shouldn't. >> >> With the word `should' I meant that it would still be possible to implement >> multiple line disciplines with the mpsafetty code. But I really think line >> disciplines should go. > > No, don't bother. Remove the entire "abstraction" of linedisciplines. > >> What should happen with the NETGRAPHDISC? > > It should be done in userland if it is necessary at all. It sounds like there is some disagreement on this point, given that that would require byte streams from serial lines to go into userspace just to be re-injected into the kernel in order to enter netgraph. Rather than do that, a more sensible route would be to look at converting ng_tty from a line discipline into a much more constrained set of I/O hooks, selected using the same ioctl but presumably a different discipline name. Unfortunately, it's not clear that snp would use the same hooks, as Netgraph presumably wants to eat the input, whereas snp just wants a copy, but perhaps there's a clean way to capture that distinction. Robert N M Watson Computer Laboratory University of Cambridge