From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 5BCA2C636D4 for ; Mon, 30 Jan 2023 19:51:16 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237175AbjA3TvO (ORCPT ); Mon, 30 Jan 2023 14:51:14 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:40894 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229972AbjA3TvJ (ORCPT ); Mon, 30 Jan 2023 14:51:09 -0500 Received: from smtp-relay-internal-0.canonical.com (smtp-relay-internal-0.canonical.com [185.125.188.122]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id E64A54617F for ; Mon, 30 Jan 2023 11:51:05 -0800 (PST) Received: from mail-wm1-f69.google.com (mail-wm1-f69.google.com [209.85.128.69]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by smtp-relay-internal-0.canonical.com (Postfix) with ESMTPS id 46D48414A7 for ; Mon, 30 Jan 2023 19:51:04 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=canonical.com; s=20210705; t=1675108264; bh=pJfLw21k4Bw+Z+qG3Fmd84/wV85iBkEmFuPWqHy43UQ=; h=Date:From:To:Cc:Subject:Message-ID:References:MIME-Version: Content-Type:In-Reply-To; b=lVcg9qM6yl+0g5aAT0mzV4lgolVh9Rkh0Z8qar2vNadro6z1/2rH/3JcpmUxG5dhh Agf0QKKBkLecZlVt+qiqrdsrGgEpdMMT25PqxqOkzw5x4Cz4Auw2f0Wgm76rNKMSF7 QIsZzu+rbW2a0lphY44BPyZ7yNuFWvYjIYlUD6zsq0iDeeNHIU18Ca8s2Y8Fr3HfYj EXx0S/tFY6YGV+vAlUHxp2ezsnTaLWCtAfoLW/mmBYy94A5jUR0MBUV3WwheLQvqfD cNHKbsTFgFzMwv8CqHVeIW8f7UU03Nvbqagmf/8HnKTkiExoh+T2/GHspke7rnud8K 0xjL4OsK+wA/g== Received: by mail-wm1-f69.google.com with SMTP id fl5-20020a05600c0b8500b003db12112fdeso7705578wmb.5 for ; Mon, 30 Jan 2023 11:51:04 -0800 (PST) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=in-reply-to:content-disposition:mime-version:references:message-id :subject:cc:to:from:date:x-gm-message-state:from:to:cc:subject:date :message-id:reply-to; bh=pJfLw21k4Bw+Z+qG3Fmd84/wV85iBkEmFuPWqHy43UQ=; b=EQ7i7ZZ57Pq8/M7ODbzMuu8ZAbbgzM/mU+kKRW7RSxqv/aNztkVB+fAg6quuWEWARG KUU7iU+J6cxAdaXC1KOHyWogjkWfs+wvw4j5ChynK3yWBDR2haJba0Iyre/0+CRU1dNh WQZ2BrmF6mVs24aVtYnFq47sPk+G/UaELO+ZpqdyV1ErNGdCdqnReH/x1zo2aqgfjOgK QM3tuUhQE+AAHfNmKwoCL+R2oRs5MubNyW+Hd1Z51cWrm5iIWeR6+PNoauuvt/EZgPE9 0ZjIBOqAh/Svk+IeEAAIYQOZMm3e8xodafrCiPOY1dwWdbA6fwNfFjxelUyhQe+EiHAN JDWg== X-Gm-Message-State: AO0yUKWabu06qqG98ZcaOH61eYmcokEDSetTvr03h4o0cecThh3RwX3G 1kpysuJ//jOxppbYvA7cCXk6VB6kkumoeDFsifDXOSxt+D8WjNRuFwE9poJfEZ1xTUr01vkcq4N JU999oBCdSW48bwkSENBm2X1mMpmotwFdVzw8pacsQA== X-Received: by 2002:a05:600c:3197:b0:3dc:496f:ad56 with SMTP id s23-20020a05600c319700b003dc496fad56mr9564113wmp.14.1675108264007; Mon, 30 Jan 2023 11:51:04 -0800 (PST) X-Google-Smtp-Source: AK7set/KQA/xahZm3/pMewnq8gulP3mMbR/RBVJSi/D+Pt02uYDk5amqQXb9r7Ir0ALl8NT2r2NktA== X-Received: by 2002:a05:600c:3197:b0:3dc:496f:ad56 with SMTP id s23-20020a05600c319700b003dc496fad56mr9564093wmp.14.1675108263766; Mon, 30 Jan 2023 11:51:03 -0800 (PST) Received: from qwirkle ([81.2.157.149]) by smtp.gmail.com with ESMTPSA id p12-20020a05600c468c00b003dc22ee5a2bsm15704537wmo.39.2023.01.30.11.51.03 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Mon, 30 Jan 2023 11:51:03 -0800 (PST) Date: Mon, 30 Jan 2023 19:51:01 +0000 From: Andrei Gherzan To: Willem de Bruijn Cc: Paolo Abeni , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Shuah Khan , netdev@vger.kernel.org, linux-kselftest@vger.kernel.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH] selftests: net: udpgso_bench_tx: Introduce exponential back-off retries Message-ID: References: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 23/01/30 06:24PM, Andrei Gherzan wrote: > On 23/01/30 12:35PM, Willem de Bruijn wrote: > > On Mon, Jan 30, 2023 at 12:31 PM Andrei Gherzan > > wrote: > > > > > > On 23/01/30 11:29AM, Willem de Bruijn wrote: > > > > On Mon, Jan 30, 2023 at 11:23 AM Andrei Gherzan > > > > wrote: > > > > > > > > > > On 23/01/30 04:15PM, Andrei Gherzan wrote: > > > > > > On 23/01/30 11:03AM, Willem de Bruijn wrote: > > > > > > > On Mon, Jan 30, 2023 at 9:28 AM Andrei Gherzan > > > > > > > wrote: > > > > > > > > > > > > > > > > On 23/01/30 08:35AM, Willem de Bruijn wrote: > > > > > > > > > On Mon, Jan 30, 2023 at 7:51 AM Andrei Gherzan > > > > > > > > > wrote: > > > > > > > > > > > > > > > > > > > > On 23/01/30 09:26AM, Paolo Abeni wrote: > > > > > > > > > > > On Fri, 2023-01-27 at 17:03 -0500, Willem de Bruijn wrote: > > > > > > > > > > > > On Fri, Jan 27, 2023 at 1:16 PM Andrei Gherzan > > > > > > > > > > > > wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > The tx and rx test programs are used in a couple of test scripts including > > > > > > > > > > > > > "udpgro_bench.sh". Taking this as an example, when the rx/tx programs > > > > > > > > > > > > > are invoked subsequently, there is a chance that the rx one is not ready to > > > > > > > > > > > > > accept socket connections. This racing bug could fail the test with at > > > > > > > > > > > > > least one of the following: > > > > > > > > > > > > > > > > > > > > > > > > > > ./udpgso_bench_tx: connect: Connection refused > > > > > > > > > > > > > ./udpgso_bench_tx: sendmsg: Connection refused > > > > > > > > > > > > > ./udpgso_bench_tx: write: Connection refused > > > > > > > > > > > > > > > > > > > > > > > > > > This change addresses this by adding routines that retry the socket > > > > > > > > > > > > > operations with an exponential back off algorithm from 100ms to 2s. > > > > > > > > > > > > > > > > > > > > > > > > > > Fixes: 3a687bef148d ("selftests: udp gso benchmark") > > > > > > > > > > > > > Signed-off-by: Andrei Gherzan > > > > > > > > > > > > > > > > > > > > > > > > Synchronizing the two processes is indeed tricky. > > > > > > > > > > > > > > > > > > > > > > > > Perhaps more robust is opening an initial TCP connection, with > > > > > > > > > > > > SO_RCVTIMEO to bound the waiting time. That covers all tests in one > > > > > > > > > > > > go. > > > > > > > > > > > > > > > > > > > > > > Another option would be waiting for the listener(tcp)/receiver(udp) > > > > > > > > > > > socket to show up in 'ss' output before firing-up the client - quite > > > > > > > > > > > alike what mptcp self-tests are doing. > > > > > > > > > > > > > > > > > > > > I like this idea. I have tested it and it works as expected with the > > > > > > > > > > exeception of: > > > > > > > > > > > > > > > > > > > > ./udpgso_bench_tx: sendmsg: No buffer space available > > > > > > > > > > > > > > > > > > > > Any ideas on how to handle this? I could retry and that works. > > > > > > > > > > > > > > > > > > This happens (also) without the zerocopy flag, right? That > > > > > > > > > > > > > > > > > > It might mean reaching the sndbuf limit, which can be adjusted with > > > > > > > > > SO_SNDBUF (or SO_SNDBUFFORCE if CAP_NET_ADMIN). Though I would not > > > > > > > > > expect this test to bump up against that limit. > > > > > > > > > > > > > > > > > > A few zerocopy specific reasons are captured in > > > > > > > > > https://www.kernel.org/doc/html/latest/networking/msg_zerocopy.html#transmission. > > > > > > > > > > > > > > > > I have dug a bit more into this, and it does look like your hint was in > > > > > > > > the right direction. The fails I'm seeing are only with the zerocopy > > > > > > > > flag. > > > > > > > > > > > > > > > > From the reasons (doc) above I can only assume optmem limit as I've > > > > > > > > reproduced it with unlimited locked pages and the fails are transient. > > > > > > > > That leaves optmem limit. Bumping the value I have by default (20480) to > > > > > > > > (2048000) made the sendmsg succeed as expected. On the other hand, the > > > > > > > > tests started to fail with something like: > > > > > > > > > > > > > > > > ./udpgso_bench_tx: Unexpected number of Zerocopy completions: 774783 > > > > > > > > expected 773707 received > > > > > > > > > > > > > > More zerocopy completions than number of sends. I have not seen this before. > > > > > > > > > > > > > > The completions are ranges of IDs, one per send call for datagram sockets. > > > > > > > > > > > > > > Even with segmentation offload, the counter increases per call, not per segment. > > > > > > > > > > > > > > Do you experience this without any other changes to udpgso_bench_tx.c. > > > > > > > Or are there perhaps additional sendmsg calls somewhere (during > > > > > > > initial sync) that are not accounted to num_sends? > > > > > > > > > > > > Indeed, that looks off. No, I have run into this without any changes in > > > > > > the tests (besides the retry routine in the shell script that waits for > > > > > > rx to come up). Also, as a data point. > > > > > > > > > > Actually wait. I don't think that is the case here. "expected" is the > > > > > number of sends. In this case we sent 1076 more messages than > > > > > completions. Am I missing something obvious? > > > > > > > > Oh indeed. > > > > > > > > Receiving fewer completions than transmission is more likely. > > > > > > Exactly, yes. > > > > > > > This should be the result of datagrams still being somewhere in the > > > > system. In a qdisc, or waiting for the network interface to return a > > > > completion notification, say. > > > > > > > > Does this remain if adding a longer wait before the final flush_errqueue? > > > > > > Yes and no. But not realiably unless I go overboard. > > > > > > > Or, really, the right fix is to keep polling there until the two are > > > > equal, up to some timeout. Currently flush_errqueue calls poll only > > > > once. > > > > > > That makes sense. I have implemented a retry and this ran for a good > > > while now. > > > > > > - flush_errqueue(fd, true); > > > + while (true) { > > > + flush_errqueue(fd, true); > > > + if ((stat_zcopies == num_sends) || (delay >= MAX_DELAY)) > > > + break; > > > + usleep(delay); > > > + delay *= 2; > > > + } > > > > > > What do you think? > > > > Thanks for running experiments. > > > > We can avoid the unconditional sleep, as the poll() inside > > flush_errqueue already takes a timeout. > > > > One option is to use start_time = clock_gettime(..) or gettimeofday > > before poll, and restart poll until either the exit condition or > > timeout is reached, with timeout = orig_time - elapsed_time. > > Yes, this was more of a quick draft. I was thinking to move it into the > flush function (while making it aware of num_sends via a parameter): > > if (do_poll) { > struct pollfd fds = {0}; > int ret; > unsigned long tnow, tstop; > > fds.fd = fd; > tnow = gettimeofday_ms(); > tstop = tnow + POLL_LOOP_TIMEOUT_MS; > while ((stat_zcopies != num_sends) && (tnow < tstop)) { > ret = poll(&fds, 1, 500); > if (ret == 0) { > if (cfg_verbose) > fprintf(stderr, "poll timeout\n"); > } else if (ret < 0) { > error(1, errno, "poll"); > } > tnow = gettimeofday_ms(); > } > } > > Does this make more sense? Obviously, this should be a do/while. Anyway, this works as expected after leaving it for a around two hours. -- Andrei Gherzan