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 X-Spam-Level: X-Spam-Status: No, score=-5.5 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_PASS,USER_AGENT_MUTT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id C1891C04EB8 for ; Fri, 30 Nov 2018 19:42:10 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 89E1B20834 for ; Fri, 30 Nov 2018 19:42:10 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 89E1B20834 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=redhat.com Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=linux-block-owner@vger.kernel.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1726159AbeLAGwa (ORCPT ); Sat, 1 Dec 2018 01:52:30 -0500 Received: from mx1.redhat.com ([209.132.183.28]:37474 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726028AbeLAGwa (ORCPT ); Sat, 1 Dec 2018 01:52:30 -0500 Received: from smtp.corp.redhat.com (int-mx07.intmail.prod.int.phx2.redhat.com [10.5.11.22]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 4E58EC02835C; Fri, 30 Nov 2018 19:42:09 +0000 (UTC) Received: from localhost (unknown [10.18.25.149]) by smtp.corp.redhat.com (Postfix) with ESMTPS id D1470104C53F; Fri, 30 Nov 2018 19:42:04 +0000 (UTC) Date: Fri, 30 Nov 2018 14:42:04 -0500 From: Mike Snitzer To: Mikulas Patocka Cc: Jens Axboe , dm-devel@redhat.com, linux-block@vger.kernel.org, "Alasdair G. Kergon" , Christoph Hellwig Subject: Re: [PATCH 1/3] block: delete part_round_stats and switch to less precise counting Message-ID: <20181130194203.GA14625@redhat.com> References: <20181128004249.147084927@debian.vm> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20181128004249.147084927@debian.vm> User-Agent: Mutt/1.5.21 (2010-09-15) X-Scanned-By: MIMEDefang 2.84 on 10.5.11.22 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.31]); Fri, 30 Nov 2018 19:42:09 +0000 (UTC) Sender: linux-block-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-block@vger.kernel.org On Tue, Nov 27 2018 at 7:42pm -0500, Mikulas Patocka wrote: > We want to convert to per-cpu in_flight counters. > > The function part_round_stats needs the in_flight counter every jiffy, it > would be too costly to sum all the percpu variables every jiffy, so it > must be deleted. part_round_stats is used to calculate two counters - > time_in_queue and io_ticks. > > time_in_queue can be calculated without part_round_stats, by adding the > duration of the I/O when the I/O ends (the value is almost as exact as the > previously calculated value, except that time for in-progress I/Os is not > counted). > > io_ticks can be approximated by increasing the value when I/O is started > or ended and the jiffies value has changed. If the I/Os take less than a > jiffy, the value is as exact as the previously calculated value. If the > I/Os take more than a jiffy, io_ticks can drift behind the previously > calculated value. > > Signed-off-by: Mikulas Patocka > > --- > block/bio.c | 26 ++++++++++++++++-- > block/blk-core.c | 64 +++------------------------------------------- > block/blk-merge.c | 1 > block/genhd.c | 4 -- > block/partition-generic.c | 4 -- > include/linux/genhd.h | 3 -- > 6 files changed, 29 insertions(+), 73 deletions(-) > > Index: linux-block/block/bio.c > =================================================================== > --- linux-block.orig/block/bio.c 2018-11-26 23:44:17.000000000 +0100 > +++ linux-block/block/bio.c 2018-11-26 23:44:17.000000000 +0100 > @@ -1663,13 +1663,29 @@ defer: > } > EXPORT_SYMBOL_GPL(bio_check_pages_dirty); > > +void update_io_ticks(int cpu, struct hd_struct *part, unsigned long now) > +{ > + unsigned long stamp; > +again: > + stamp = READ_ONCE(part->stamp); > + if (unlikely(stamp != now)) { > + if (likely(cmpxchg(&part->stamp, stamp, now) == stamp)) { > + __part_stat_add(cpu, part, io_ticks, 1); > + } > + } > + if (part->partno) { > + part = &part_to_disk(part)->part0; > + goto again; > + } > +} > + > void generic_start_io_acct(struct request_queue *q, int op, > unsigned long sectors, struct hd_struct *part) > { > const int sgrp = op_stat_group(op); > int cpu = part_stat_lock(); > > - part_round_stats(q, cpu, part); > + update_io_ticks(cpu, part, jiffies); > part_stat_inc(cpu, part, ios[sgrp]); > part_stat_add(cpu, part, sectors[sgrp], sectors); > part_inc_in_flight(q, part, op_is_write(op)); > @@ -1681,12 +1697,16 @@ EXPORT_SYMBOL(generic_start_io_acct); > void generic_end_io_acct(struct request_queue *q, int req_op, > struct hd_struct *part, unsigned long start_time) > { > - unsigned long duration = jiffies - start_time; > + unsigned long now = jiffies; > + unsigned long duration = now - start_time; > const int sgrp = op_stat_group(req_op); > int cpu = part_stat_lock(); > > + update_io_ticks(cpu, part, now); > part_stat_add(cpu, part, nsecs[sgrp], jiffies_to_nsecs(duration)); > - part_round_stats(q, cpu, part); > + part_stat_add(cpu, part, time_in_queue, duration); > + if (part->partno) > + part_stat_add(cpu, &part_to_disk(part)->part0, time_in_queue, duration); > part_dec_in_flight(q, part, op_is_write(req_op)); > > part_stat_unlock(); > Index: linux-block/block/blk-core.c > =================================================================== > --- linux-block.orig/block/blk-core.c 2018-11-26 23:44:17.000000000 +0100 > +++ linux-block/block/blk-core.c 2018-11-26 23:44:17.000000000 +0100 > @@ -583,63 +583,6 @@ struct request *blk_get_request(struct r > } > EXPORT_SYMBOL(blk_get_request); > > -static void part_round_stats_single(struct request_queue *q, int cpu, > - struct hd_struct *part, unsigned long now, > - unsigned int inflight) > -{ > - if (inflight) { > - __part_stat_add(cpu, part, time_in_queue, > - inflight * (now - part->stamp)); > - __part_stat_add(cpu, part, io_ticks, (now - part->stamp)); > - } > - part->stamp = now; > -} > - > -/** > - * part_round_stats() - Round off the performance stats on a struct disk_stats. > - * @q: target block queue > - * @cpu: cpu number for stats access > - * @part: target partition > - * > - * The average IO queue length and utilisation statistics are maintained > - * by observing the current state of the queue length and the amount of > - * time it has been in this state for. > - * > - * Normally, that accounting is done on IO completion, but that can result > - * in more than a second's worth of IO being accounted for within any one > - * second, leading to >100% utilisation. To deal with that, we call this > - * function to do a round-off before returning the results when reading > - * /proc/diskstats. This accounts immediately for all queue usage up to > - * the current jiffies and restarts the counters again. > - */ > -void part_round_stats(struct request_queue *q, int cpu, struct hd_struct *part) > -{ > - struct hd_struct *part2 = NULL; > - unsigned long now = jiffies; > - unsigned int inflight[2]; > - int stats = 0; > - > - if (part->stamp != now) > - stats |= 1; > - > - if (part->partno) { > - part2 = &part_to_disk(part)->part0; > - if (part2->stamp != now) > - stats |= 2; > - } > - > - if (!stats) > - return; > - > - part_in_flight(q, part, inflight); > - > - if (stats & 2) > - part_round_stats_single(q, cpu, part2, now, inflight[1]); > - if (stats & 1) > - part_round_stats_single(q, cpu, part, now, inflight[0]); > -} > -EXPORT_SYMBOL_GPL(part_round_stats); > - > void blk_put_request(struct request *req) > { > blk_mq_free_request(req); > @@ -1408,9 +1351,11 @@ void blk_account_io_done(struct request > cpu = part_stat_lock(); > part = req->part; > > + update_io_ticks(cpu, part, jiffies); > part_stat_inc(cpu, part, ios[sgrp]); > part_stat_add(cpu, part, nsecs[sgrp], now - req->start_time_ns); > - part_round_stats(req->q, cpu, part); > + part_stat_add(cpu, part, time_in_queue, nsecs_to_jiffies64(now - req->start_time_ns)); (Extra indentation offered a clue...) Are you missing an 'if (part->partno)' conditional here? > + part_stat_add(cpu, &part_to_disk(part)->part0, time_in_queue, nsecs_to_jiffies64(now - req->start_time_ns)); > part_dec_in_flight(req->q, part, rq_data_dir(req)); > > hd_struct_put(part);