From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1761212AbXJYELw (ORCPT ); Thu, 25 Oct 2007 00:11:52 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752083AbXJYELo (ORCPT ); Thu, 25 Oct 2007 00:11:44 -0400 Received: from smtp2.linux-foundation.org ([207.189.120.14]:33637 "EHLO smtp2.linux-foundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751192AbXJYELn (ORCPT ); Thu, 25 Oct 2007 00:11:43 -0400 Date: Wed, 24 Oct 2007 21:11:40 -0700 From: Andrew Morton To: Matthew Wilcox Cc: torvalds@linux-foundation.org, linux-kernel@vger.kernel.org, Matthew Wilcox Subject: Re: [PATCH 4/5] Add lock_page_killable Message-Id: <20071024211140.329fdd90.akpm@linux-foundation.org> In-Reply-To: <11932286982653-git-send-email-matthew@wil.cx> References: <11932286982175-git-send-email-matthew@wil.cx> <11932286982653-git-send-email-matthew@wil.cx> X-Mailer: Sylpheed 2.4.1 (GTK+ 2.8.17; x86_64-unknown-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 24 Oct 2007 08:24:57 -0400 Matthew Wilcox wrote: > and associated infrastructure such as sync_page_killable and > fatal_signal_pending. Use lock_page_killable in do_generic_mapping_read() > to allow us to kill `cat' of a file on an NFS-mounted filesystem. whoa, big change. What exactly are the semantics here? If the process has actually been killed (ie: we know that userspace won't be running again) then we break out of a lock_page() and allow the process to exit? ie: it's basically invisible to userspace? If so, it sounds OK. I guess. We're still screwed if the process is doing a synchronous write and lots of other scenarios. How well has this been tested? Have the NFS guys had a think about it? Why does it return -EIO from read() and not -EINTR?