From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758354Ab3CUODl (ORCPT ); Thu, 21 Mar 2013 10:03:41 -0400 Received: from relay.parallels.com ([195.214.232.42]:38909 "EHLO relay.parallels.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757474Ab3CUODk (ORCPT ); Thu, 21 Mar 2013 10:03:40 -0400 Subject: [PATCH 0/4] fuse: fix accounting background requests (v2) To: miklos@szeredi.hu From: "Maxim V. Patlasov" Cc: fuse-devel@lists.sourceforge.net, xemul@parallels.com, linux-kernel@vger.kernel.org, devel@openvz.org, dev@parallels.com Date: Thu, 21 Mar 2013 18:01:53 +0400 Message-ID: <20130321140047.4051.6701.stgit@maximpc.sw.ru> User-Agent: StGit/0.15 MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi, The feature was added long time ago (commit 08a53cdc...) with the comment: > A task may have at most one synchronous request allocated. So these requests > need not be otherwise limited. > > However the number of background requests (release, forget, asynchronous > reads, interrupted requests) can grow indefinitely. This can be used by a > malicous user to cause FUSE to allocate arbitrary amounts of unswappable > kernel memory, denying service. > > For this reason add a limit for the number of background requests, and block > allocations of new requests until the number goes bellow the limit. However, the implementation suffers from the following problems: 1. Latency of synchronous requests. As soon as fc->num_background hits the limit, all allocations are blocked: both for synchronous and background requests. This is unnecessary - as the comment cited above states, synchronous requests need not be limited (by fuse). Moreover, sometimes it's very inconvenient. For example, a dozen of tasks aggressively writing to mmap()-ed area may block 'ls' for long while (>1min in my experiments). 2. Thundering herd problem. When fc->num_background falls below the limit, request_end() calls wake_up_all(&fc->blocked_waitq). This wakes up all waiters while it's not impossible that the first waiter getting new request will immediately put it to background increasing fc->num_background again. (experimenting with mmap()-ed writes I observed 2x slowdown as compared with fuse after applying this patch-set) The patch-set re-works fuse_get_req (and its callers) to throttle only requests intended for background processing. Having this done, it becomes possible to use exclusive wakeups in chained manner: request_end() wakes up a waiter, the waiter allocates new request and submits it for background processing, the processing ends in request_end() where another wakeup happens an so on. Changed in v2: - rebased on for-next branch of the fuse tree - fixed race when processing request begins before init-reply came Thanks, Maxim --- Maxim V. Patlasov (4): fuse: make request allocations for background processing explicit fuse: add flag fc->uninitialized fuse: skip blocking on allocations of synchronous requests fuse: implement exclusive wakeup for blocked_waitq fs/fuse/cuse.c | 3 ++ fs/fuse/dev.c | 69 +++++++++++++++++++++++++++++++++++++++++++----------- fs/fuse/file.c | 6 +++-- fs/fuse/fuse_i.h | 8 ++++++ fs/fuse/inode.c | 4 +++ 5 files changed, 73 insertions(+), 17 deletions(-) -- Signature