From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755355AbdKAUx5 (ORCPT ); Wed, 1 Nov 2017 16:53:57 -0400 Received: from mail.linuxfoundation.org ([140.211.169.12]:36918 "EHLO mail.linuxfoundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755260AbdKAUx4 (ORCPT ); Wed, 1 Nov 2017 16:53:56 -0400 Date: Wed, 1 Nov 2017 13:53:54 -0700 From: Andrew Morton To: Davidlohr Bueso Cc: Jason Baron , linux-kernel@vger.kernel.org, Alexander Viro , Salman Qazi , Hou Tao Subject: Re: [PATCH] epoll: remove ep_call_nested() from ep_eventpoll_poll() Message-Id: <20171101135354.1ffec54e1fdedcca0f2086ce@linux-foundation.org> In-Reply-To: <20171031145821.ieflhlvs3la25vdw@linux-n805> References: <1509430214-5599-1-git-send-email-jbaron@akamai.com> <20171031145821.ieflhlvs3la25vdw@linux-n805> X-Mailer: Sylpheed 3.4.1 (GTK+ 2.24.23; x86_64-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, 31 Oct 2017 07:58:21 -0700 Davidlohr Bueso wrote: > On Tue, 31 Oct 2017, Jason Baron wrote: > > >The use of ep_call_nested() in ep_eventpoll_poll(), which is the .poll > >routine for an epoll fd, is used to prevent excessively deep epoll > >nesting, and to prevent circular paths. However, we are already preventing > >these conditions during EPOLL_CTL_ADD. In terms of too deep epoll chains, > >we do in fact allow deep nesting of the epoll fds themselves (deeper > >than EP_MAX_NESTS), however we don't allow more than EP_MAX_NESTS when > >an epoll file descriptor is actually connected to a wakeup source. Thus, > >we do not require the use of ep_call_nested(), since ep_eventpoll_poll(), > >which is called via ep_scan_ready_list() only continues nesting if there > >are events available. Since ep_call_nested() is implemented using a global > >lock, applications that make use of nested epoll can see large performance > >improvements with this change. > > Improvements are quite obscene actually, such as for the following epoll_wait() > benchmark with 2 level nesting on a 80 core IvyBridge: > > ncpus vanilla dirty delta > 1 2447092 3028315 +23.75% > 4 231265 2986954 +1191.57% > 8 121631 2898796 +2283.27% > 16 59749 2902056 +4757.07% > 32 26837 2326314 +8568.30% > 64 12926 1341281 +10276.61% > > (http://linux-scalability.org/epoll/epoll-test.c) This is tempting, but boy it is late in the -rc cycle. How important are these workloads? Would the world end if we held off on this for 4.15?