From nobody Tue Sep 16 21:33:18 2025 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 797DBC53210 for ; Tue, 3 Jan 2023 14:55:35 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237476AbjACOze (ORCPT ); Tue, 3 Jan 2023 09:55:34 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:60718 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237937AbjACOz2 (ORCPT ); Tue, 3 Jan 2023 09:55:28 -0500 Received: from mail-ej1-x632.google.com (mail-ej1-x632.google.com [IPv6:2a00:1450:4864:20::632]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 20119FD34 for ; Tue, 3 Jan 2023 06:55:22 -0800 (PST) Received: by mail-ej1-x632.google.com with SMTP id m18so74169185eji.5 for ; Tue, 03 Jan 2023 06:55:22 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=7cSfPTrMY4yBJKJIutaXLPmmK6MgiHJjY1AusI2IKUY=; b=w31v1xN1n/GWe/n2j6YhPIB6TOC5Ddw32ewUAR6D8F+LpDFfpGc9oUQk5sjitr55ys xrL1Y337Xv+ZhcyR/8bOeYoaFNVpPNKj8J7EUzjmF5L7gaF6gXVXjwm3p0FjgVKSNgDt 8pnDKZ7B2TWPgvSDHONHaxZwOk8jxGtmYIOBCPurJLlOUCZrBnsQzSTWqDwIyZg6WXo1 fTVOtjsOxt3AjLu8V35CfrwwAugwLTJ3dLL3fE3iAI8lrtInokx1f5bKWQd3oEzpFM83 zaHNzPXU5715eiGNNKyfGhE6ErkNyc34Tfp/yXoWRZ7DJ75yTJ1IUdqsiI8G+iDY0Mef 4kLQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=7cSfPTrMY4yBJKJIutaXLPmmK6MgiHJjY1AusI2IKUY=; b=UzJT7RtgoqTIWb5iry+zVZLR/sDPILprvhymTnM3uD3cIerbyyRDkTMKGG+Fml6yK7 6RcEjSMUxsQPPCELsiRNznXon7bu2xfkT+pTuyGFUwaOpiKUq8XJgR9NRgTBb/Sonk10 hajFZi/qi7ETcxw++ZwhUYGuDJihgsxEdzylWkbHNN45syvv9q/IrdsWB95saBz7FyxU 6LjyCIM+WwVhIWwjgMBch580PEU7sBH8ymwkDdpMTKf1RGxwXoGUkaPRVvKkO03kO7xq F/p0cLSONkWiuOjim+oyUUDVDsluARtR4gn8BTuUv8eCTePcgaIV3Qsbebu71TjvNDpy Ecvw== X-Gm-Message-State: AFqh2ko97HLvzVousBC87JNwWqZMM/DEERjUqdFJg3fbK9F9O8m6ifJT E1K9gpkLWT1QAaTk2W2SrigNrA== X-Google-Smtp-Source: AMrXdXtHyCdDXDY3tr2zxU5DiCgtOEBRo4qL8Sw2WAg+0GABU6PesVJaSq7cYh656EAG+uPRaZlo+Q== X-Received: by 2002:a17:907:7707:b0:844:c651:ce4b with SMTP id kw7-20020a170907770700b00844c651ce4bmr35937037ejc.33.1672757720540; Tue, 03 Jan 2023 06:55:20 -0800 (PST) Received: from localhost.localdomain (mob-5-91-46-2.net.vodafone.it. [5.91.46.2]) by smtp.gmail.com with ESMTPSA id g17-20020a17090604d100b007c16fdc93ddsm14122595eja.81.2023.01.03.06.55.19 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Tue, 03 Jan 2023 06:55:20 -0800 (PST) From: Paolo Valente To: Jens Axboe Cc: linux-block@vger.kernel.org, linux-kernel@vger.kernel.org, arie.vanderhoeven@seagate.com, rory.c.chen@seagate.com, glen.valante@linaro.org, damien.lemoal@opensource.wdc.com, Paolo Valente , Gabriele Felici , Carmine Zaccagnino Subject: [PATCH V13 REBASED 1/8] block, bfq: split sync bfq_queues on a per-actuator basis Date: Tue, 3 Jan 2023 15:54:56 +0100 Message-Id: <20230103145503.71712-2-paolo.valente@linaro.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20230103145503.71712-1-paolo.valente@linaro.org> References: <20230103145503.71712-1-paolo.valente@linaro.org> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Single-LUN multi-actuator SCSI drives, as well as all multi-actuator SATA drives appear as a single device to the I/O subsystem [1]. Yet they address commands to different actuators internally, as a function of Logical Block Addressing (LBAs). A given sector is reachable by only one of the actuators. For example, Seagate=E2=80=99s Serial Advanced Technology Attachment (SATA) version contains two actuators and maps the lower half of the SATA LBA space to the lower actuator and the upper half to the upper actuator. Evidently, to fully utilize actuators, no actuator must be left idle or underutilized while there is pending I/O for it. The block layer must somehow control the load of each actuator individually. This commit lays the ground for allowing BFQ to provide such a per-actuator control. BFQ associates an I/O-request sync bfq_queue with each process doing synchronous I/O, or with a group of processes, in case of queue merging. Then BFQ serves one bfq_queue at a time. While in service, a bfq_queue is emptied in request-position order. Yet the same process, or group of processes, may generate I/O for different actuators. In this case, different streams of I/O (each for a different actuator) get all inserted into the same sync bfq_queue. So there is basically no individual control on when each stream is served, i.e., on when the I/O requests of the stream are picked from the bfq_queue and dispatched to the drive. This commit enables BFQ to control the service of each actuator individually for synchronous I/O, by simply splitting each sync bfq_queue into N queues, one for each actuator. In other words, a sync bfq_queue is now associated to a pair (process, actuator). As a consequence of this split, the per-queue proportional-share policy implemented by BFQ will guarantee that the sync I/O generated for each actuator, by each process, receives its fair share of service. This is just a preparatory patch. If the I/O of the same process happens to be sent to different queues, then each of these queues may undergo queue merging. To handle this event, the bfq_io_cq data structure must be properly extended. In addition, stable merging must be disabled to avoid loss of control on individual actuators. Finally, also async queues must be split. These issues are described in detail and addressed in next commits. As for this commit, although multiple per-process bfq_queues are provided, the I/O of each process or group of processes is still sent to only one queue, regardless of the actuator the I/O is for. The forwarding to distinct bfq_queues will be enabled after addressing the above issues. [1] https://www.linaro.org/blog/budget-fair-queueing-bfq-linux-io-scheduler= -optimizations-for-multi-actuator-sata-hard-drives/ Reviewed-by: Damien Le Moal Signed-off-by: Gabriele Felici Signed-off-by: Carmine Zaccagnino Signed-off-by: Paolo Valente --- block/bfq-cgroup.c | 91 +++++++++++++------------ block/bfq-iosched.c | 160 +++++++++++++++++++++++++++++--------------- block/bfq-iosched.h | 51 +++++++++++--- 3 files changed, 195 insertions(+), 107 deletions(-) diff --git a/block/bfq-cgroup.c b/block/bfq-cgroup.c index 1b2829e99dad..bd7bf0f8121d 100644 --- a/block/bfq-cgroup.c +++ b/block/bfq-cgroup.c @@ -714,6 +714,46 @@ void bfq_bfqq_move(struct bfq_data *bfqd, struct bfq_q= ueue *bfqq, bfq_put_queue(bfqq); } =20 +static void bfq_sync_bfqq_move(struct bfq_data *bfqd, + struct bfq_queue *sync_bfqq, + struct bfq_io_cq *bic, + struct bfq_group *bfqg, + unsigned int act_idx) +{ + struct bfq_queue *bfqq; + + if (!sync_bfqq->new_bfqq && !bfq_bfqq_coop(sync_bfqq)) { + /* We are the only user of this bfqq, just move it */ + if (sync_bfqq->entity.sched_data !=3D &bfqg->sched_data) + bfq_bfqq_move(bfqd, sync_bfqq, bfqg); + return; + } + + /* + * The queue was merged to a different queue. Check + * that the merge chain still belongs to the same + * cgroup. + */ + for (bfqq =3D sync_bfqq; bfqq; bfqq =3D bfqq->new_bfqq) + if (bfqq->entity.sched_data !=3D &bfqg->sched_data) + break; + if (bfqq) { + /* + * Some queue changed cgroup so the merge is not valid + * anymore. We cannot easily just cancel the merge (by + * clearing new_bfqq) as there may be other processes + * using this queue and holding refs to all queues + * below sync_bfqq->new_bfqq. Similarly if the merge + * already happened, we need to detach from bfqq now + * so that we cannot merge bio to a request from the + * old cgroup. + */ + bfq_put_cooperator(sync_bfqq); + bfq_release_process_ref(bfqd, sync_bfqq); + bic_set_bfqq(bic, NULL, true, act_idx); + } +} + /** * __bfq_bic_change_cgroup - move @bic to @bfqg. * @bfqd: the queue descriptor. @@ -728,53 +768,20 @@ static void __bfq_bic_change_cgroup(struct bfq_data *= bfqd, struct bfq_io_cq *bic, struct bfq_group *bfqg) { - struct bfq_queue *async_bfqq =3D bic_to_bfqq(bic, false); - struct bfq_queue *sync_bfqq =3D bic_to_bfqq(bic, true); - struct bfq_entity *entity; + unsigned int act_idx; =20 - if (async_bfqq) { - entity =3D &async_bfqq->entity; + for (act_idx =3D 0; act_idx < bfqd->num_actuators; act_idx++) { + struct bfq_queue *async_bfqq =3D bic_to_bfqq(bic, false, act_idx); + struct bfq_queue *sync_bfqq =3D bic_to_bfqq(bic, true, act_idx); =20 - if (entity->sched_data !=3D &bfqg->sched_data) { - bic_set_bfqq(bic, NULL, false); + if (async_bfqq && + async_bfqq->entity.sched_data !=3D &bfqg->sched_data) { + bic_set_bfqq(bic, NULL, false, act_idx); bfq_release_process_ref(bfqd, async_bfqq); } - } =20 - if (sync_bfqq) { - if (!sync_bfqq->new_bfqq && !bfq_bfqq_coop(sync_bfqq)) { - /* We are the only user of this bfqq, just move it */ - if (sync_bfqq->entity.sched_data !=3D &bfqg->sched_data) - bfq_bfqq_move(bfqd, sync_bfqq, bfqg); - } else { - struct bfq_queue *bfqq; - - /* - * The queue was merged to a different queue. Check - * that the merge chain still belongs to the same - * cgroup. - */ - for (bfqq =3D sync_bfqq; bfqq; bfqq =3D bfqq->new_bfqq) - if (bfqq->entity.sched_data !=3D - &bfqg->sched_data) - break; - if (bfqq) { - /* - * Some queue changed cgroup so the merge is - * not valid anymore. We cannot easily just - * cancel the merge (by clearing new_bfqq) as - * there may be other processes using this - * queue and holding refs to all queues below - * sync_bfqq->new_bfqq. Similarly if the merge - * already happened, we need to detach from - * bfqq now so that we cannot merge bio to a - * request from the old cgroup. - */ - bfq_put_cooperator(sync_bfqq); - bfq_release_process_ref(bfqd, sync_bfqq); - bic_set_bfqq(bic, NULL, true); - } - } + if (sync_bfqq) + bfq_sync_bfqq_move(bfqd, sync_bfqq, bic, bfqg, act_idx); } } =20 diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index ccf2204477a5..597f5a109092 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -377,16 +377,23 @@ static const unsigned long bfq_late_stable_merging = =3D 600; #define RQ_BIC(rq) ((struct bfq_io_cq *)((rq)->elv.priv[0])) #define RQ_BFQQ(rq) ((rq)->elv.priv[1]) =20 -struct bfq_queue *bic_to_bfqq(struct bfq_io_cq *bic, bool is_sync) +struct bfq_queue *bic_to_bfqq(struct bfq_io_cq *bic, bool is_sync, + unsigned int actuator_idx) { - return bic->bfqq[is_sync]; + if (is_sync) + return bic->bfqq[1][actuator_idx]; + + return bic->bfqq[0][actuator_idx]; } =20 static void bfq_put_stable_ref(struct bfq_queue *bfqq); =20 -void bic_set_bfqq(struct bfq_io_cq *bic, struct bfq_queue *bfqq, bool is_s= ync) +void bic_set_bfqq(struct bfq_io_cq *bic, + struct bfq_queue *bfqq, + bool is_sync, + unsigned int actuator_idx) { - struct bfq_queue *old_bfqq =3D bic->bfqq[is_sync]; + struct bfq_queue *old_bfqq =3D bic->bfqq[is_sync][actuator_idx]; =20 /* Clear bic pointer if bfqq is detached from this bic */ if (old_bfqq && old_bfqq->bic =3D=3D bic) @@ -405,7 +412,10 @@ void bic_set_bfqq(struct bfq_io_cq *bic, struct bfq_qu= eue *bfqq, bool is_sync) * we cancel the stable merge if * bic->stable_merge_bfqq =3D=3D bfqq. */ - bic->bfqq[is_sync] =3D bfqq; + if (is_sync) + bic->bfqq[1][actuator_idx] =3D bfqq; + else + bic->bfqq[0][actuator_idx] =3D bfqq; =20 if (bfqq && bic->stable_merge_bfqq =3D=3D bfqq) { /* @@ -678,9 +688,9 @@ static void bfq_limit_depth(blk_opf_t opf, struct blk_m= q_alloc_data *data) { struct bfq_data *bfqd =3D data->q->elevator->elevator_data; struct bfq_io_cq *bic =3D bfq_bic_lookup(data->q); - struct bfq_queue *bfqq =3D bic ? bic_to_bfqq(bic, op_is_sync(opf)) : NULL; int depth; unsigned limit =3D data->q->nr_requests; + unsigned int act_idx; =20 /* Sync reads have full depth available */ if (op_is_sync(opf) && !op_is_write(opf)) { @@ -690,14 +700,21 @@ static void bfq_limit_depth(blk_opf_t opf, struct blk= _mq_alloc_data *data) limit =3D (limit * depth) >> bfqd->full_depth_shift; } =20 - /* - * Does queue (or any parent entity) exceed number of requests that - * should be available to it? Heavily limit depth so that it cannot - * consume more available requests and thus starve other entities. - */ - if (bfqq && bfqq_request_over_limit(bfqq, limit)) - depth =3D 1; + for (act_idx =3D 0; bic && act_idx < bfqd->num_actuators; act_idx++) { + struct bfq_queue *bfqq =3D + bic_to_bfqq(bic, op_is_sync(opf), act_idx); =20 + /* + * Does queue (or any parent entity) exceed number of + * requests that should be available to it? Heavily + * limit depth so that it cannot consume more + * available requests and thus starve other entities. + */ + if (bfqq && bfqq_request_over_limit(bfqq, limit)) { + depth =3D 1; + break; + } + } bfq_log(bfqd, "[%s] wr_busy %d sync %d depth %u", __func__, bfqd->wr_busy_queues, op_is_sync(opf), depth); if (depth) @@ -1766,6 +1783,18 @@ static bool bfq_bfqq_higher_class_or_weight(struct b= fq_queue *bfqq, return bfqq_weight > in_serv_weight; } =20 +/* + * Get the index of the actuator that will serve bio. + */ +static unsigned int bfq_actuator_index(struct bfq_data *bfqd, struct bio *= bio) +{ + /* + * Multi-actuator support not complete yet, so always return 0 + * for the moment (to keep incomplete mechanisms off). + */ + return 0; +} + static bool bfq_better_to_idle(struct bfq_queue *bfqq); =20 static void bfq_bfqq_handle_idle_busy_switch(struct bfq_data *bfqd, @@ -2098,7 +2127,7 @@ static void bfq_check_waker(struct bfq_data *bfqd, st= ruct bfq_queue *bfqq, * We reset waker detection logic also if too much time has passed * since the first detection. If wakeups are rare, pointless idling * doesn't hurt throughput that much. The condition below makes sure - * we do not uselessly idle blocking waker in more than 1/64 cases.=20 + * we do not uselessly idle blocking waker in more than 1/64 cases. */ if (bfqd->last_completed_rq_bfqq !=3D bfqq->tentative_waker_bfqq || @@ -2418,7 +2447,8 @@ static bool bfq_bio_merge(struct request_queue *q, st= ruct bio *bio, */ bfq_bic_update_cgroup(bic, bio); =20 - bfqd->bio_bfqq =3D bic_to_bfqq(bic, op_is_sync(bio->bi_opf)); + bfqd->bio_bfqq =3D bic_to_bfqq(bic, op_is_sync(bio->bi_opf), + bfq_actuator_index(bfqd, bio)); } else { bfqd->bio_bfqq =3D NULL; } @@ -3114,7 +3144,7 @@ bfq_merge_bfqqs(struct bfq_data *bfqd, struct bfq_io_= cq *bic, /* * Merge queues (that is, let bic redirect its requests to new_bfqq) */ - bic_set_bfqq(bic, new_bfqq, true); + bic_set_bfqq(bic, new_bfqq, true, bfqq->actuator_idx); bfq_mark_bfqq_coop(new_bfqq); /* * new_bfqq now belongs to at least two bics (it is a shared queue): @@ -4748,11 +4778,8 @@ static struct bfq_queue *bfq_select_queue(struct bfq= _data *bfqd) */ if (bfq_bfqq_wait_request(bfqq) || (bfqq->dispatched !=3D 0 && bfq_better_to_idle(bfqq))) { - struct bfq_queue *async_bfqq =3D - bfqq->bic && bfqq->bic->bfqq[0] && - bfq_bfqq_busy(bfqq->bic->bfqq[0]) && - bfqq->bic->bfqq[0]->next_rq ? - bfqq->bic->bfqq[0] : NULL; + unsigned int act_idx =3D bfqq->actuator_idx; + struct bfq_queue *async_bfqq =3D NULL; struct bfq_queue *blocked_bfqq =3D !hlist_empty(&bfqq->woken_list) ? container_of(bfqq->woken_list.first, @@ -4760,6 +4787,10 @@ static struct bfq_queue *bfq_select_queue(struct bfq= _data *bfqd) woken_list_node) : NULL; =20 + if (bfqq->bic && bfqq->bic->bfqq[0][act_idx] && + bfq_bfqq_busy(bfqq->bic->bfqq[0][act_idx]) && + bfqq->bic->bfqq[0][act_idx]->next_rq) + async_bfqq =3D bfqq->bic->bfqq[0][act_idx]; /* * The next four mutually-exclusive ifs decide * whether to try injection, and choose the queue to @@ -4844,7 +4875,7 @@ static struct bfq_queue *bfq_select_queue(struct bfq_= data *bfqd) icq_to_bic(async_bfqq->next_rq->elv.icq) =3D=3D bfqq->bic && bfq_serv_to_charge(async_bfqq->next_rq, async_bfqq) <=3D bfq_bfqq_budget_left(async_bfqq)) - bfqq =3D bfqq->bic->bfqq[0]; + bfqq =3D bfqq->bic->bfqq[0][act_idx]; else if (bfqq->waker_bfqq && bfq_bfqq_busy(bfqq->waker_bfqq) && bfqq->waker_bfqq->next_rq && @@ -5305,48 +5336,54 @@ static void bfq_exit_bfqq(struct bfq_data *bfqd, st= ruct bfq_queue *bfqq) bfq_release_process_ref(bfqd, bfqq); } =20 -static void bfq_exit_icq_bfqq(struct bfq_io_cq *bic, bool is_sync) +static void bfq_exit_icq_bfqq(struct bfq_io_cq *bic, bool is_sync, + unsigned int actuator_idx) { - struct bfq_queue *bfqq =3D bic_to_bfqq(bic, is_sync); + struct bfq_queue *bfqq =3D bic_to_bfqq(bic, is_sync, actuator_idx); struct bfq_data *bfqd; =20 if (bfqq) bfqd =3D bfqq->bfqd; /* NULL if scheduler already exited */ =20 if (bfqq && bfqd) { - unsigned long flags; - - spin_lock_irqsave(&bfqd->lock, flags); - bic_set_bfqq(bic, NULL, is_sync); + bic_set_bfqq(bic, NULL, is_sync, actuator_idx); bfq_exit_bfqq(bfqd, bfqq); - spin_unlock_irqrestore(&bfqd->lock, flags); } } =20 static void bfq_exit_icq(struct io_cq *icq) { struct bfq_io_cq *bic =3D icq_to_bic(icq); + struct bfq_data *bfqd =3D bic_to_bfqd(bic); + unsigned long flags; + unsigned int act_idx; + /* + * If bfqd and thus bfqd->num_actuators is not available any + * longer, then cycle over all possible per-actuator bfqqs in + * next loop. We rely on bic being zeroed on creation, and + * therefore on its unused per-actuator fields being NULL. + */ + unsigned int num_actuators =3D BFQ_MAX_ACTUATORS; =20 - if (bic->stable_merge_bfqq) { - struct bfq_data *bfqd =3D bic->stable_merge_bfqq->bfqd; + /* + * bfqd is NULL if scheduler already exited, and in that case + * this is the last time these queues are accessed. + */ + if (bfqd) { + spin_lock_irqsave(&bfqd->lock, flags); + num_actuators =3D bfqd->num_actuators; + } =20 - /* - * bfqd is NULL if scheduler already exited, and in - * that case this is the last time bfqq is accessed. - */ - if (bfqd) { - unsigned long flags; + if (bic->stable_merge_bfqq) + bfq_put_stable_ref(bic->stable_merge_bfqq); =20 - spin_lock_irqsave(&bfqd->lock, flags); - bfq_put_stable_ref(bic->stable_merge_bfqq); - spin_unlock_irqrestore(&bfqd->lock, flags); - } else { - bfq_put_stable_ref(bic->stable_merge_bfqq); - } + for (act_idx =3D 0; act_idx < num_actuators; act_idx++) { + bfq_exit_icq_bfqq(bic, true, act_idx); + bfq_exit_icq_bfqq(bic, false, act_idx); } =20 - bfq_exit_icq_bfqq(bic, true); - bfq_exit_icq_bfqq(bic, false); + if (bfqd) + spin_unlock_irqrestore(&bfqd->lock, flags); } =20 /* @@ -5423,23 +5460,25 @@ static void bfq_check_ioprio_change(struct bfq_io_c= q *bic, struct bio *bio) =20 bic->ioprio =3D ioprio; =20 - bfqq =3D bic_to_bfqq(bic, false); + bfqq =3D bic_to_bfqq(bic, false, bfq_actuator_index(bfqd, bio)); if (bfqq) { bfq_release_process_ref(bfqd, bfqq); bfqq =3D bfq_get_queue(bfqd, bio, false, bic, true); - bic_set_bfqq(bic, bfqq, false); + bic_set_bfqq(bic, bfqq, false, bfq_actuator_index(bfqd, bio)); } =20 - bfqq =3D bic_to_bfqq(bic, true); + bfqq =3D bic_to_bfqq(bic, true, bfq_actuator_index(bfqd, bio)); if (bfqq) bfq_set_next_ioprio_data(bfqq, bic); } =20 static void bfq_init_bfqq(struct bfq_data *bfqd, struct bfq_queue *bfqq, - struct bfq_io_cq *bic, pid_t pid, int is_sync) + struct bfq_io_cq *bic, pid_t pid, int is_sync, + unsigned int act_idx) { u64 now_ns =3D ktime_get_ns(); =20 + bfqq->actuator_idx =3D act_idx; RB_CLEAR_NODE(&bfqq->entity.rb_node); INIT_LIST_HEAD(&bfqq->fifo); INIT_HLIST_NODE(&bfqq->burst_list_node); @@ -5692,7 +5731,7 @@ static struct bfq_queue *bfq_get_queue(struct bfq_dat= a *bfqd, =20 if (bfqq) { bfq_init_bfqq(bfqd, bfqq, bic, current->pid, - is_sync); + is_sync, bfq_actuator_index(bfqd, bio)); bfq_init_entity(&bfqq->entity, bfqg); bfq_log_bfqq(bfqd, bfqq, "allocated"); } else { @@ -6007,7 +6046,8 @@ static bool __bfq_insert_request(struct bfq_data *bfq= d, struct request *rq) * then complete the merge and redirect it to * new_bfqq. */ - if (bic_to_bfqq(RQ_BIC(rq), 1) =3D=3D bfqq) + if (bic_to_bfqq(RQ_BIC(rq), true, + bfq_actuator_index(bfqd, rq->bio)) =3D=3D bfqq) bfq_merge_bfqqs(bfqd, RQ_BIC(rq), bfqq, new_bfqq); =20 @@ -6562,7 +6602,7 @@ bfq_split_bfqq(struct bfq_io_cq *bic, struct bfq_queu= e *bfqq) return bfqq; } =20 - bic_set_bfqq(bic, NULL, true); + bic_set_bfqq(bic, NULL, true, bfqq->actuator_idx); =20 bfq_put_cooperator(bfqq); =20 @@ -6576,7 +6616,8 @@ static struct bfq_queue *bfq_get_bfqq_handle_split(st= ruct bfq_data *bfqd, bool split, bool is_sync, bool *new_queue) { - struct bfq_queue *bfqq =3D bic_to_bfqq(bic, is_sync); + unsigned int act_idx =3D bfq_actuator_index(bfqd, bio); + struct bfq_queue *bfqq =3D bic_to_bfqq(bic, is_sync, act_idx); =20 if (likely(bfqq && bfqq !=3D &bfqd->oom_bfqq)) return bfqq; @@ -6588,7 +6629,7 @@ static struct bfq_queue *bfq_get_bfqq_handle_split(st= ruct bfq_data *bfqd, bfq_put_queue(bfqq); bfqq =3D bfq_get_queue(bfqd, bio, is_sync, bic, split); =20 - bic_set_bfqq(bic, bfqq, is_sync); + bic_set_bfqq(bic, bfqq, is_sync, act_idx); if (split && is_sync) { if ((bic->was_in_burst_list && bfqd->large_burst) || bic->saved_in_large_burst) @@ -7036,8 +7077,10 @@ static int bfq_init_queue(struct request_queue *q, s= truct elevator_type *e) * Our fallback bfqq if bfq_find_alloc_queue() runs into OOM issues. * Grab a permanent reference to it, so that the normal code flow * will not attempt to free it. + * Set zero as actuator index: we will pretend that + * all I/O requests are for the same actuator. */ - bfq_init_bfqq(bfqd, &bfqd->oom_bfqq, NULL, 1, 0); + bfq_init_bfqq(bfqd, &bfqd->oom_bfqq, NULL, 1, 0, 0); bfqd->oom_bfqq.ref++; bfqd->oom_bfqq.new_ioprio =3D BFQ_DEFAULT_QUEUE_IOPRIO; bfqd->oom_bfqq.new_ioprio_class =3D IOPRIO_CLASS_BE; @@ -7056,6 +7099,13 @@ static int bfq_init_queue(struct request_queue *q, s= truct elevator_type *e) =20 bfqd->queue =3D q; =20 + /* + * Multi-actuator support not complete yet, unconditionally + * set to only one actuator for the moment (to keep incomplete + * mechanisms off). + */ + bfqd->num_actuators =3D 1; + INIT_LIST_HEAD(&bfqd->dispatch); =20 hrtimer_init(&bfqd->idle_slice_timer, CLOCK_MONOTONIC, diff --git a/block/bfq-iosched.h b/block/bfq-iosched.h index 41aa151ccc22..3c6b21b2cd95 100644 --- a/block/bfq-iosched.h +++ b/block/bfq-iosched.h @@ -33,6 +33,14 @@ */ #define BFQ_SOFTRT_WEIGHT_FACTOR 100 =20 +/* + * Maximum number of actuators supported. This constant is used simply + * to define the size of the static array that will contain + * per-actuator data. The current value is hopefully a good upper + * bound to the possible number of actuators of any actual drive. + */ +#define BFQ_MAX_ACTUATORS 8 + struct bfq_entity; =20 /** @@ -227,12 +235,14 @@ struct bfq_ttime { * struct bfq_queue - leaf schedulable entity. * * A bfq_queue is a leaf request queue; it can be associated with an - * io_context or more, if it is async or shared between cooperating - * processes. @cgroup holds a reference to the cgroup, to be sure that it - * does not disappear while a bfqq still references it (mostly to avoid - * races between request issuing and task migration followed by cgroup - * destruction). - * All the fields are protected by the queue lock of the containing bfqd. + * io_context or more, if it is async or shared between cooperating + * processes. Besides, it contains I/O requests for only one actuator + * (an io_context is associated with a different bfq_queue for each + * actuator it generates I/O for). @cgroup holds a reference to the + * cgroup, to be sure that it does not disappear while a bfqq still + * references it (mostly to avoid races between request issuing and + * task migration followed by cgroup destruction). All the fields are + * protected by the queue lock of the containing bfqd. */ struct bfq_queue { /* reference counter */ @@ -397,6 +407,9 @@ struct bfq_queue { * the woken queues when this queue exits. */ struct hlist_head woken_list; + + /* index of the actuator this queue is associated with */ + unsigned int actuator_idx; }; =20 /** @@ -405,8 +418,17 @@ struct bfq_queue { struct bfq_io_cq { /* associated io_cq structure */ struct io_cq icq; /* must be the first member */ - /* array of two process queues, the sync and the async */ - struct bfq_queue *bfqq[2]; + /* + * Matrix of associated process queues: first row for async + * queues, second row sync queues. Each row contains one + * column for each actuator. An I/O request generated by the + * process is inserted into the queue pointed by bfqq[i][j] if + * the request is to be served by the j-th actuator of the + * drive, where i=3D=3D0 or i=3D=3D1, depending on whether the request + * is async or sync. So there is a distinct queue for each + * actuator. + */ + struct bfq_queue *bfqq[2][BFQ_MAX_ACTUATORS]; /* per (request_queue, blkcg) ioprio */ int ioprio; #ifdef CONFIG_BFQ_GROUP_IOSCHED @@ -772,6 +794,13 @@ struct bfq_data { */ unsigned int word_depths[2][2]; unsigned int full_depth_shift; + + /* + * Number of independent actuators. This is equal to 1 in + * case of single-actuator drives. + */ + unsigned int num_actuators; + }; =20 enum bfqq_state_flags { @@ -969,8 +998,10 @@ struct bfq_group { =20 extern const int bfq_timeout; =20 -struct bfq_queue *bic_to_bfqq(struct bfq_io_cq *bic, bool is_sync); -void bic_set_bfqq(struct bfq_io_cq *bic, struct bfq_queue *bfqq, bool is_s= ync); +struct bfq_queue *bic_to_bfqq(struct bfq_io_cq *bic, bool is_sync, + unsigned int actuator_idx); +void bic_set_bfqq(struct bfq_io_cq *bic, struct bfq_queue *bfqq, bool is_s= ync, + unsigned int actuator_idx); struct bfq_data *bic_to_bfqd(struct bfq_io_cq *bic); void bfq_pos_tree_add_move(struct bfq_data *bfqd, struct bfq_queue *bfqq); void bfq_weights_tree_add(struct bfq_queue *bfqq); --=20 2.20.1 From nobody Tue Sep 16 21:33:18 2025 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 85C94C3DA7D for ; Tue, 3 Jan 2023 14:55:39 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237873AbjACOzg (ORCPT ); Tue, 3 Jan 2023 09:55:36 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:60514 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230459AbjACOz2 (ORCPT ); Tue, 3 Jan 2023 09:55:28 -0500 Received: from mail-ej1-x634.google.com (mail-ej1-x634.google.com [IPv6:2a00:1450:4864:20::634]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4EA7BFD0C for ; Tue, 3 Jan 2023 06:55:23 -0800 (PST) Received: by mail-ej1-x634.google.com with SMTP id u19so74243726ejm.8 for ; Tue, 03 Jan 2023 06:55:23 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=d0d4MosATHMJxt0bdurE2AL3gK0xP2e0HXm81ECVDQs=; b=nIZbEVpZFdPxZXFuW1jTgdH1T6VdOM/cbMLygn3bm/yx7vDj8tm97hV+ZWCWhJoYhp S2xlNJdiUN+tK2W2rpZ5gRBobG7o+tTZkrF8AbtGH61whnjmXLigjlJ4JVbbOixFpCTR 7lt0i+dGNWFWCXwsPBgwgBouBwHZ/g13bCxPpq2kk8kTAfoYerWIiZii9MqOI3ELNcN6 7jjwhM5Cx0IfjMG2rSavglggjlOQez8zQYtd3vPAl/hHvKDGiwtyTET7NlI14rk3RYyR T4VO/TloMajCikv84fxlbEx1+bK7YkxlWcfyqoqgwN2uvpD2+2LhWt4LlqDytBsXS387 ANMQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=d0d4MosATHMJxt0bdurE2AL3gK0xP2e0HXm81ECVDQs=; b=7eWN2/XUMsXz+hSWVyifGYOeCGgagtu9D+EcKWuLwVN8HCoBelEErGkirDxqyEZcUn 7N93yB/UFbhLMWkqbhPRmDYkzC6K++xZKHFGTa9NOoeYLtVq8jWtjL0JTxokclf1+APQ 1zoo7FAninPVKHI2wZbYRAvWKKzVdeBJHcdQYo/MQIdlOvEzGzArDR0UQBKZi5WKVeaN +YAr1RB2cSmHatCsWuRfQdQpYFcdTNwLusUTrz/anBMD/K6De11O7YBjOcznoJeXpAHu tv9d6Jqsdxh4m0+Wm6fi1MNMeQUXzwGRHLMNhkFGYwZBp5YICimth1aE2O6ZqkbUZ/S/ QY7A== X-Gm-Message-State: AFqh2kqCWrVOmGYxOaEbJX3WcsR0kTvrAtRDZd5gpgGvW+PN7YOTEqyE 6bV450Th1qyXL/nBzolKrqnm5w== X-Google-Smtp-Source: AMrXdXtKiz1+4PfEbvupQ8hd59N0IyFXZHSvBkbroZ+zohW8twn8CF9iTWP2YmKDQm0VSHveNsraqQ== X-Received: by 2002:a17:906:308b:b0:7ae:cda1:76d0 with SMTP id 11-20020a170906308b00b007aecda176d0mr36632976ejv.15.1672757721870; Tue, 03 Jan 2023 06:55:21 -0800 (PST) Received: from localhost.localdomain (mob-5-91-46-2.net.vodafone.it. [5.91.46.2]) by smtp.gmail.com with ESMTPSA id g17-20020a17090604d100b007c16fdc93ddsm14122595eja.81.2023.01.03.06.55.20 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Tue, 03 Jan 2023 06:55:21 -0800 (PST) From: Paolo Valente To: Jens Axboe Cc: linux-block@vger.kernel.org, linux-kernel@vger.kernel.org, arie.vanderhoeven@seagate.com, rory.c.chen@seagate.com, glen.valante@linaro.org, damien.lemoal@opensource.wdc.com, Paolo Valente Subject: [PATCH V13 REBASED 2/8] block, bfq: forbid stable merging of queues associated with different actuators Date: Tue, 3 Jan 2023 15:54:57 +0100 Message-Id: <20230103145503.71712-3-paolo.valente@linaro.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20230103145503.71712-1-paolo.valente@linaro.org> References: <20230103145503.71712-1-paolo.valente@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" If queues associated with different actuators are merged, then control is lost on each actuator. Therefore some actuator may be underutilized, and throughput may decrease. This problem cannot occur with basic queue merging, because the latter is triggered by spatial locality, and sectors for different actuators are not close to each other. Yet it may happen with stable merging. To address this issue, this commit prevents stable merging from occurring among queues associated with different actuators. Reviewed-by: Damien Le Moal Signed-off-by: Paolo Valente --- block/bfq-iosched.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index 597f5a109092..b21ca9111b87 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -5647,9 +5647,13 @@ static struct bfq_queue *bfq_do_or_sched_stable_merg= e(struct bfq_data *bfqd, * it has been set already, but too long ago, then move it * forward to bfqq. Finally, move also if bfqq belongs to a * different group than last_bfqq_created, or if bfqq has a - * different ioprio or ioprio_class. If none of these - * conditions holds true, then try an early stable merge or - * schedule a delayed stable merge. + * different ioprio, ioprio_class or actuator_idx. If none of + * these conditions holds true, then try an early stable merge + * or schedule a delayed stable merge. As for the condition on + * actuator_idx, the reason is that, if queues associated with + * different actuators are merged, then control is lost on + * each actuator. Therefore some actuator may be + * underutilized, and throughput may decrease. * * A delayed merge is scheduled (instead of performing an * early merge), in case bfqq might soon prove to be more @@ -5667,7 +5671,8 @@ static struct bfq_queue *bfq_do_or_sched_stable_merge= (struct bfq_data *bfqd, bfqq->creation_time) || bfqq->entity.parent !=3D last_bfqq_created->entity.parent || bfqq->ioprio !=3D last_bfqq_created->ioprio || - bfqq->ioprio_class !=3D last_bfqq_created->ioprio_class) + bfqq->ioprio_class !=3D last_bfqq_created->ioprio_class || + bfqq->actuator_idx !=3D last_bfqq_created->actuator_idx) *source_bfqq =3D bfqq; else if (time_after_eq(last_bfqq_created->creation_time + bfqd->bfq_burst_interval, --=20 2.20.1 From nobody Tue Sep 16 21:33:18 2025 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 CED3BC46467 for ; Tue, 3 Jan 2023 14:55:48 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S238044AbjACOzr (ORCPT ); Tue, 3 Jan 2023 09:55:47 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:60612 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237967AbjACOz3 (ORCPT ); Tue, 3 Jan 2023 09:55:29 -0500 Received: from mail-ej1-x636.google.com (mail-ej1-x636.google.com [IPv6:2a00:1450:4864:20::636]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id DC53FFD29 for ; Tue, 3 Jan 2023 06:55:23 -0800 (PST) Received: by mail-ej1-x636.google.com with SMTP id fy8so10336136ejc.13 for ; Tue, 03 Jan 2023 06:55:23 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=dFHGav7fAtKcn2krHx/valIkDrVr74B/YppQ0nSTImY=; b=uyowHYrS7um+ZA2V9bfPnWWpOpRXDAsJVuAa9TBW8ir7/WqTOcihDxGv1Iihoc7b8i 1ywg+TiKZFAfh0KVAHUVDTNiCbqAQr/sNasdN5PXab4p4vitvvHa8/PecUMPgsf2P3hV vRcLSnKyNNRfoKXw9FYpLL53jaeHPx8pNprdWhCEUwapJnYF4so+YGVO+QIMZ1of0CiN /hOHVGNBOZcEvvD3LRJt16KKPeqC71e7RVAke5P/Toi7zm+EINsanEciYSybZBKl6zox yi+4/WOwF/FLwlQM7c8gv6r1Qr1VVJGqysB702IK3EdenZPVigXHBgCJihWwc8+fYfBO ou8Q== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=dFHGav7fAtKcn2krHx/valIkDrVr74B/YppQ0nSTImY=; b=e3DN33GL4bmY4ZU7zuXyqK1dABgVAaeQ57JoX4wVc68pBtAJ6Mh+2noYlS5qDqU8lv LRhQgGjRLdojA2Ee8NUTgIIMO4eeH7P5TYOxxyUqv8+YMgAXcRNoYactPF7u8DBOOepz ctLDpSNlieFEW0yMujj3mM05e4pSsxw0nQF13QD+T5PvBtJMWRdu4U8gYniDh7ceEXTh aa4AzcENiXb0rElVFcbTKoYv04JFT8/sWVr0qXv4gtcxuOF7CjmjWGrYAzQTWOVPy+zr 1gerjYe+rXeUW7MS5MuNe3b00BGW3jLUIMFTzfkDlEVvjFOSXJXlI8PcQrm6hB6q3HbN jaAQ== X-Gm-Message-State: AFqh2kpTssHvJHHufLLhZlrgvkobQoLFol6MbcJyqLCnq1WUm5M4juj5 cDwMDfskKfD/eJKrVhQQ/hRUTA== X-Google-Smtp-Source: AMrXdXv2NUhIUatNI+cYt9YjMq400HLXzVHo2y2uhodkhlAlb8jVWv0yYLHMGTn9q+NZYopjo0kdVQ== X-Received: by 2002:a17:907:a07c:b0:841:13b0:7238 with SMTP id ia28-20020a170907a07c00b0084113b07238mr36738970ejc.25.1672757723326; Tue, 03 Jan 2023 06:55:23 -0800 (PST) Received: from localhost.localdomain (mob-5-91-46-2.net.vodafone.it. [5.91.46.2]) by smtp.gmail.com with ESMTPSA id g17-20020a17090604d100b007c16fdc93ddsm14122595eja.81.2023.01.03.06.55.21 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Tue, 03 Jan 2023 06:55:22 -0800 (PST) From: Paolo Valente To: Jens Axboe Cc: linux-block@vger.kernel.org, linux-kernel@vger.kernel.org, arie.vanderhoeven@seagate.com, rory.c.chen@seagate.com, glen.valante@linaro.org, damien.lemoal@opensource.wdc.com, Paolo Valente , Gianmarco Lusvardi , Giulio Barabino , Emiliano Maccaferri Subject: [PATCH V13 REBASED 3/8] block, bfq: move io_cq-persistent bfqq data into a dedicated struct Date: Tue, 3 Jan 2023 15:54:58 +0100 Message-Id: <20230103145503.71712-4-paolo.valente@linaro.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20230103145503.71712-1-paolo.valente@linaro.org> References: <20230103145503.71712-1-paolo.valente@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" With a multi-actuator drive, a process may get associated with multiple bfq_queues: one queue for each of the N actuators. So, the bfq_io_cq data structure must be able to accommodate its per-queue persistent information for N queues. Currently it stores this information for just one queue, in several scalar fields. This is a preparatory commit for moving to accommodating persistent information for N queues. In particular, this commit packs all the above scalar fields into a single data structure. Then there is now only one field, in bfq_io_cq, that stores all the above information. This scalar field will then be turned into an array by a following commit. Suggested-by: Damien Le Moal Reviewed-by: Damien Le Moal Signed-off-by: Gianmarco Lusvardi Signed-off-by: Giulio Barabino Signed-off-by: Emiliano Maccaferri Signed-off-by: Paolo Valente --- block/bfq-iosched.c | 138 +++++++++++++++++++++++++------------------- block/bfq-iosched.h | 52 ++++++++++------- 2 files changed, 110 insertions(+), 80 deletions(-) diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index b21ca9111b87..cefd16b9cbd1 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -395,10 +395,6 @@ void bic_set_bfqq(struct bfq_io_cq *bic, { struct bfq_queue *old_bfqq =3D bic->bfqq[is_sync][actuator_idx]; =20 - /* Clear bic pointer if bfqq is detached from this bic */ - if (old_bfqq && old_bfqq->bic =3D=3D bic) - old_bfqq->bic =3D NULL; - /* * If bfqq !=3D NULL, then a non-stable queue merge between * bic->bfqq and bfqq is happening here. This causes troubles @@ -412,12 +408,18 @@ void bic_set_bfqq(struct bfq_io_cq *bic, * we cancel the stable merge if * bic->stable_merge_bfqq =3D=3D bfqq. */ + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; + + /* Clear bic pointer if bfqq is detached from this bic */ + if (old_bfqq && old_bfqq->bic =3D=3D bic) + old_bfqq->bic =3D NULL; + if (is_sync) bic->bfqq[1][actuator_idx] =3D bfqq; else bic->bfqq[0][actuator_idx] =3D bfqq; =20 - if (bfqq && bic->stable_merge_bfqq =3D=3D bfqq) { + if (bfqq && bfqq_data->stable_merge_bfqq =3D=3D bfqq) { /* * Actually, these same instructions are executed also * in bfq_setup_cooperator, in case of abort or actual @@ -426,9 +428,9 @@ void bic_set_bfqq(struct bfq_io_cq *bic, * did so, we would nest even more complexity in this * function. */ - bfq_put_stable_ref(bic->stable_merge_bfqq); + bfq_put_stable_ref(bfqq_data->stable_merge_bfqq); =20 - bic->stable_merge_bfqq =3D NULL; + bfqq_data->stable_merge_bfqq =3D NULL; } } =20 @@ -1133,38 +1135,40 @@ static void bfq_bfqq_resume_state(struct bfq_queue *bfqq, struct bfq_data *bfqd, struct bfq_io_cq *bic, bool bfq_already_existing) { + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; unsigned int old_wr_coeff =3D 1; bool busy =3D bfq_already_existing && bfq_bfqq_busy(bfqq); =20 - if (bic->saved_has_short_ttime) + if (bfqq_data->saved_has_short_ttime) bfq_mark_bfqq_has_short_ttime(bfqq); else bfq_clear_bfqq_has_short_ttime(bfqq); =20 - if (bic->saved_IO_bound) + if (bfqq_data->saved_IO_bound) bfq_mark_bfqq_IO_bound(bfqq); else bfq_clear_bfqq_IO_bound(bfqq); =20 - bfqq->last_serv_time_ns =3D bic->saved_last_serv_time_ns; - bfqq->inject_limit =3D bic->saved_inject_limit; - bfqq->decrease_time_jif =3D bic->saved_decrease_time_jif; + bfqq->last_serv_time_ns =3D bfqq_data->saved_last_serv_time_ns; + bfqq->inject_limit =3D bfqq_data->saved_inject_limit; + bfqq->decrease_time_jif =3D bfqq_data->saved_decrease_time_jif; =20 - bfqq->entity.new_weight =3D bic->saved_weight; - bfqq->ttime =3D bic->saved_ttime; - bfqq->io_start_time =3D bic->saved_io_start_time; - bfqq->tot_idle_time =3D bic->saved_tot_idle_time; + bfqq->entity.new_weight =3D bfqq_data->saved_weight; + bfqq->ttime =3D bfqq_data->saved_ttime; + bfqq->io_start_time =3D bfqq_data->saved_io_start_time; + bfqq->tot_idle_time =3D bfqq_data->saved_tot_idle_time; /* * Restore weight coefficient only if low_latency is on */ if (bfqd->low_latency) { old_wr_coeff =3D bfqq->wr_coeff; - bfqq->wr_coeff =3D bic->saved_wr_coeff; + bfqq->wr_coeff =3D bfqq_data->saved_wr_coeff; } - bfqq->service_from_wr =3D bic->saved_service_from_wr; - bfqq->wr_start_at_switch_to_srt =3D bic->saved_wr_start_at_switch_to_srt; - bfqq->last_wr_start_finish =3D bic->saved_last_wr_start_finish; - bfqq->wr_cur_max_time =3D bic->saved_wr_cur_max_time; + bfqq->service_from_wr =3D bfqq_data->saved_service_from_wr; + bfqq->wr_start_at_switch_to_srt =3D + bfqq_data->saved_wr_start_at_switch_to_srt; + bfqq->last_wr_start_finish =3D bfqq_data->saved_last_wr_start_finish; + bfqq->wr_cur_max_time =3D bfqq_data->saved_wr_cur_max_time; =20 if (bfqq->wr_coeff > 1 && (bfq_bfqq_in_large_burst(bfqq) || time_is_before_jiffies(bfqq->last_wr_start_finish + @@ -1849,7 +1853,7 @@ static void bfq_bfqq_handle_idle_busy_switch(struct b= fq_data *bfqd, wr_or_deserves_wr =3D bfqd->low_latency && (bfqq->wr_coeff > 1 || (bfq_bfqq_sync(bfqq) && - (bfqq->bic || RQ_BIC(rq)->stably_merged) && + (bfqq->bic || RQ_BIC(rq)->bfqq_data.stably_merged) && (*interactive || soft_rt))); =20 /* @@ -2849,6 +2853,7 @@ bfq_setup_cooperator(struct bfq_data *bfqd, struct bf= q_queue *bfqq, void *io_struct, bool request, struct bfq_io_cq *bic) { struct bfq_queue *in_service_bfqq, *new_bfqq; + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; =20 /* if a merge has already been setup, then proceed with that first */ if (bfqq->new_bfqq) @@ -2870,21 +2875,21 @@ bfq_setup_cooperator(struct bfq_data *bfqd, struct = bfq_queue *bfqq, * stable merging) also if bic is associated with a * sync queue, but this bfqq is async */ - if (bfq_bfqq_sync(bfqq) && bic->stable_merge_bfqq && + if (bfq_bfqq_sync(bfqq) && bfqq_data->stable_merge_bfqq && !bfq_bfqq_just_created(bfqq) && time_is_before_jiffies(bfqq->split_time + msecs_to_jiffies(bfq_late_stable_merging)) && time_is_before_jiffies(bfqq->creation_time + msecs_to_jiffies(bfq_late_stable_merging))) { struct bfq_queue *stable_merge_bfqq =3D - bic->stable_merge_bfqq; + bfqq_data->stable_merge_bfqq; int proc_ref =3D min(bfqq_process_refs(bfqq), bfqq_process_refs(stable_merge_bfqq)); =20 /* deschedule stable merge, because done or aborted here */ bfq_put_stable_ref(stable_merge_bfqq); =20 - bic->stable_merge_bfqq =3D NULL; + bfqq_data->stable_merge_bfqq =3D NULL; =20 if (!idling_boosts_thr_without_issues(bfqd, bfqq) && proc_ref > 0) { @@ -2893,10 +2898,10 @@ bfq_setup_cooperator(struct bfq_data *bfqd, struct = bfq_queue *bfqq, bfq_setup_merge(bfqq, stable_merge_bfqq); =20 if (new_bfqq) { - bic->stably_merged =3D true; + bfqq_data->stably_merged =3D true; if (new_bfqq->bic) - new_bfqq->bic->stably_merged =3D - true; + new_bfqq->bic->bfqq_data.stably_merged =3D + true; } return new_bfqq; } else @@ -2995,6 +3000,7 @@ bfq_setup_cooperator(struct bfq_data *bfqd, struct bf= q_queue *bfqq, static void bfq_bfqq_save_state(struct bfq_queue *bfqq) { struct bfq_io_cq *bic =3D bfqq->bic; + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; =20 /* * If !bfqq->bic, the queue is already shared or its requests @@ -3004,18 +3010,21 @@ static void bfq_bfqq_save_state(struct bfq_queue *b= fqq) if (!bic) return; =20 - bic->saved_last_serv_time_ns =3D bfqq->last_serv_time_ns; - bic->saved_inject_limit =3D bfqq->inject_limit; - bic->saved_decrease_time_jif =3D bfqq->decrease_time_jif; - - bic->saved_weight =3D bfqq->entity.orig_weight; - bic->saved_ttime =3D bfqq->ttime; - bic->saved_has_short_ttime =3D bfq_bfqq_has_short_ttime(bfqq); - bic->saved_IO_bound =3D bfq_bfqq_IO_bound(bfqq); - bic->saved_io_start_time =3D bfqq->io_start_time; - bic->saved_tot_idle_time =3D bfqq->tot_idle_time; - bic->saved_in_large_burst =3D bfq_bfqq_in_large_burst(bfqq); - bic->was_in_burst_list =3D !hlist_unhashed(&bfqq->burst_list_node); + bfqq_data->saved_last_serv_time_ns =3D bfqq->last_serv_time_ns; + bfqq_data->saved_inject_limit =3D bfqq->inject_limit; + bfqq_data->saved_decrease_time_jif =3D bfqq->decrease_time_jif; + + bfqq_data->saved_weight =3D bfqq->entity.orig_weight; + bfqq_data->saved_ttime =3D bfqq->ttime; + bfqq_data->saved_has_short_ttime =3D + bfq_bfqq_has_short_ttime(bfqq); + bfqq_data->saved_IO_bound =3D bfq_bfqq_IO_bound(bfqq); + bfqq_data->saved_io_start_time =3D bfqq->io_start_time; + bfqq_data->saved_tot_idle_time =3D bfqq->tot_idle_time; + bfqq_data->saved_in_large_burst =3D bfq_bfqq_in_large_burst(bfqq); + bfqq_data->was_in_burst_list =3D + !hlist_unhashed(&bfqq->burst_list_node); + if (unlikely(bfq_bfqq_just_created(bfqq) && !bfq_bfqq_in_large_burst(bfqq) && bfqq->bfqd->low_latency)) { @@ -3028,17 +3037,21 @@ static void bfq_bfqq_save_state(struct bfq_queue *b= fqq) * to bfqq, so that to avoid that bfqq unjustly fails * to enjoy weight raising if split soon. */ - bic->saved_wr_coeff =3D bfqq->bfqd->bfq_wr_coeff; - bic->saved_wr_start_at_switch_to_srt =3D bfq_smallest_from_now(); - bic->saved_wr_cur_max_time =3D bfq_wr_duration(bfqq->bfqd); - bic->saved_last_wr_start_finish =3D jiffies; + bfqq_data->saved_wr_coeff =3D bfqq->bfqd->bfq_wr_coeff; + bfqq_data->saved_wr_start_at_switch_to_srt =3D + bfq_smallest_from_now(); + bfqq_data->saved_wr_cur_max_time =3D + bfq_wr_duration(bfqq->bfqd); + bfqq_data->saved_last_wr_start_finish =3D jiffies; } else { - bic->saved_wr_coeff =3D bfqq->wr_coeff; - bic->saved_wr_start_at_switch_to_srt =3D + bfqq_data->saved_wr_coeff =3D bfqq->wr_coeff; + bfqq_data->saved_wr_start_at_switch_to_srt =3D bfqq->wr_start_at_switch_to_srt; - bic->saved_service_from_wr =3D bfqq->service_from_wr; - bic->saved_last_wr_start_finish =3D bfqq->last_wr_start_finish; - bic->saved_wr_cur_max_time =3D bfqq->wr_cur_max_time; + bfqq_data->saved_service_from_wr =3D + bfqq->service_from_wr; + bfqq_data->saved_last_wr_start_finish =3D + bfqq->last_wr_start_finish; + bfqq_data->saved_wr_cur_max_time =3D bfqq->wr_cur_max_time; } } =20 @@ -5364,6 +5377,7 @@ static void bfq_exit_icq(struct io_cq *icq) * therefore on its unused per-actuator fields being NULL. */ unsigned int num_actuators =3D BFQ_MAX_ACTUATORS; + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; =20 /* * bfqd is NULL if scheduler already exited, and in that case @@ -5374,8 +5388,8 @@ static void bfq_exit_icq(struct io_cq *icq) num_actuators =3D bfqd->num_actuators; } =20 - if (bic->stable_merge_bfqq) - bfq_put_stable_ref(bic->stable_merge_bfqq); + if (bfqq_data->stable_merge_bfqq) + bfq_put_stable_ref(bfqq_data->stable_merge_bfqq); =20 for (act_idx =3D 0; act_idx < num_actuators; act_idx++) { bfq_exit_icq_bfqq(bic, true, act_idx); @@ -5566,13 +5580,14 @@ bfq_do_early_stable_merge(struct bfq_data *bfqd, st= ruct bfq_queue *bfqq, { struct bfq_queue *new_bfqq =3D bfq_setup_merge(bfqq, last_bfqq_created); + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; =20 if (!new_bfqq) return bfqq; =20 if (new_bfqq->bic) - new_bfqq->bic->stably_merged =3D true; - bic->stably_merged =3D true; + new_bfqq->bic->bfqq_data.stably_merged =3D true; + bfqq_data->stably_merged =3D true; =20 /* * Reusing merge functions. This implies that @@ -5641,6 +5656,7 @@ static struct bfq_queue *bfq_do_or_sched_stable_merge= (struct bfq_data *bfqd, &bfqd->last_bfqq_created; =20 struct bfq_queue *last_bfqq_created =3D *source_bfqq; + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; =20 /* * If last_bfqq_created has not been set yet, then init it. If @@ -5702,7 +5718,7 @@ static struct bfq_queue *bfq_do_or_sched_stable_merge= (struct bfq_data *bfqd, /* * Record the bfqq to merge to. */ - bic->stable_merge_bfqq =3D last_bfqq_created; + bfqq_data->stable_merge_bfqq =3D last_bfqq_created; } } =20 @@ -6623,6 +6639,7 @@ static struct bfq_queue *bfq_get_bfqq_handle_split(st= ruct bfq_data *bfqd, { unsigned int act_idx =3D bfq_actuator_index(bfqd, bio); struct bfq_queue *bfqq =3D bic_to_bfqq(bic, is_sync, act_idx); + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; =20 if (likely(bfqq && bfqq !=3D &bfqd->oom_bfqq)) return bfqq; @@ -6636,12 +6653,12 @@ static struct bfq_queue *bfq_get_bfqq_handle_split(= struct bfq_data *bfqd, =20 bic_set_bfqq(bic, bfqq, is_sync, act_idx); if (split && is_sync) { - if ((bic->was_in_burst_list && bfqd->large_burst) || - bic->saved_in_large_burst) + if ((bfqq_data->was_in_burst_list && bfqd->large_burst) || + bfqq_data->saved_in_large_burst) bfq_mark_bfqq_in_large_burst(bfqq); else { bfq_clear_bfqq_in_large_burst(bfqq); - if (bic->was_in_burst_list) + if (bfqq_data->was_in_burst_list) /* * If bfqq was in the current * burst list before being @@ -6730,6 +6747,7 @@ static struct bfq_queue *bfq_init_rq(struct request *= rq) struct bfq_queue *bfqq; bool new_queue =3D false; bool bfqq_already_existing =3D false, split =3D false; + struct bfq_iocq_bfqq_data *bfqq_data; =20 if (unlikely(!rq->elv.icq)) return NULL; @@ -6753,15 +6771,17 @@ static struct bfq_queue *bfq_init_rq(struct request= *rq) bfqq =3D bfq_get_bfqq_handle_split(bfqd, bic, bio, false, is_sync, &new_queue); =20 + bfqq_data =3D &bic->bfqq_data; + if (likely(!new_queue)) { /* If the queue was seeky for too long, break it apart. */ if (bfq_bfqq_coop(bfqq) && bfq_bfqq_split_coop(bfqq) && - !bic->stably_merged) { + !bfqq_data->stably_merged) { struct bfq_queue *old_bfqq =3D bfqq; =20 /* Update bic before losing reference to bfqq */ if (bfq_bfqq_in_large_burst(bfqq)) - bic->saved_in_large_burst =3D true; + bfqq_data->saved_in_large_burst =3D true; =20 bfqq =3D bfq_split_bfqq(bic, bfqq); split =3D true; diff --git a/block/bfq-iosched.h b/block/bfq-iosched.h index 3c6b21b2cd95..8e0215b8cc30 100644 --- a/block/bfq-iosched.h +++ b/block/bfq-iosched.h @@ -413,27 +413,9 @@ struct bfq_queue { }; =20 /** - * struct bfq_io_cq - per (request_queue, io_context) structure. - */ -struct bfq_io_cq { - /* associated io_cq structure */ - struct io_cq icq; /* must be the first member */ - /* - * Matrix of associated process queues: first row for async - * queues, second row sync queues. Each row contains one - * column for each actuator. An I/O request generated by the - * process is inserted into the queue pointed by bfqq[i][j] if - * the request is to be served by the j-th actuator of the - * drive, where i=3D=3D0 or i=3D=3D1, depending on whether the request - * is async or sync. So there is a distinct queue for each - * actuator. - */ - struct bfq_queue *bfqq[2][BFQ_MAX_ACTUATORS]; - /* per (request_queue, blkcg) ioprio */ - int ioprio; -#ifdef CONFIG_BFQ_GROUP_IOSCHED - uint64_t blkcg_serial_nr; /* the current blkcg serial */ -#endif +* struct bfq_data - bfqq data unique and persistent for associated bfq_io_= cq +*/ +struct bfq_iocq_bfqq_data { /* * Snapshot of the has_short_time flag before merging; taken * to remember its value while the queue is merged, so as to @@ -488,6 +470,34 @@ struct bfq_io_cq { struct bfq_queue *stable_merge_bfqq; =20 bool stably_merged; /* non splittable if true */ +}; + +/** + * struct bfq_io_cq - per (request_queue, io_context) structure. + */ +struct bfq_io_cq { + /* associated io_cq structure */ + struct io_cq icq; /* must be the first member */ + /* + * Matrix of associated process queues: first row for async + * queues, second row sync queues. Each row contains one + * column for each actuator. An I/O request generated by the + * process is inserted into the queue pointed by bfqq[i][j] if + * the request is to be served by the j-th actuator of the + * drive, where i=3D=3D0 or i=3D=3D1, depending on whether the request + * is async or sync. So there is a distinct queue for each + * actuator. + */ + struct bfq_queue *bfqq[2][BFQ_MAX_ACTUATORS]; + /* per (request_queue, blkcg) ioprio */ + int ioprio; +#ifdef CONFIG_BFQ_GROUP_IOSCHED + uint64_t blkcg_serial_nr; /* the current blkcg serial */ +#endif + + /* persistent data for associated synchronous process queue */ + struct bfq_iocq_bfqq_data bfqq_data; + unsigned int requests; /* Number of requests this process has in flight */ }; =20 --=20 2.20.1 From nobody Tue Sep 16 21:33:18 2025 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 A0669C54EBC for ; Tue, 3 Jan 2023 14:55:54 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S238077AbjACOzw (ORCPT ); Tue, 3 Jan 2023 09:55:52 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:60664 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237822AbjACOza (ORCPT ); Tue, 3 Jan 2023 09:55:30 -0500 Received: from mail-ed1-x530.google.com (mail-ed1-x530.google.com [IPv6:2a00:1450:4864:20::530]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 9CF4811A3B for ; Tue, 3 Jan 2023 06:55:26 -0800 (PST) Received: by mail-ed1-x530.google.com with SMTP id b88so36840375edf.6 for ; Tue, 03 Jan 2023 06:55:26 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=gzJ9bA+w5xTwtmeG3+bNpReH2QJDYQbMcs0UfcoR8Rw=; b=XyBWHBEPviYdMqrMfCXvIMy+0V2kVoLtKsAJrt3g+11AlyfrasLQQtetd3jxUkjQVB +Cqn3Iq5OYbgP16jBP+o/pNwKX1psLnLXxcaBz4zD9IIyFmKrl0w5S/kVwmMQjJrtt8v HgJBkj9NkxyO5/yK+MPdCAAsf0PzSVHkU80MMji9z8dB1WvkkHsuI+CcszeJRPLCcuR5 c3Rh1q0mr6Y6sdeSlKrphdJmTf0JWb001sLx/CkYcZFYln4MMv1xEn/fmjqzB31vpN7d Vi9ovckUEWves4pq75wHub/IP9i+VyUmrgR+SHYLOrCCimp6+5RJMJr5fLd9k4zkrZQP hYFQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=gzJ9bA+w5xTwtmeG3+bNpReH2QJDYQbMcs0UfcoR8Rw=; b=aaE+cge32yDlp/9IncKtxcRy4EeSrvLFycgHniV0aw8sXfrRuDazrzYk/Pbsnth9bp gTVvMopqchds41y3rGbP9fxr20W/zCE2Me2PkQkPi4/qq6nPpRA5pQjc4xK2FNBhdS+k 2DDpLkWS0R+yUrtostgOCJA6gdaOmXGMYMODj6nogkFnGwvnCxWojrDm30wgR84zZbVY rxRjdgLNruIUmw+lyS2fJBS0PmLF69+ZjGPf5vg2tK35AaEPzz3c/ksu7ySbuGwYGHVc 2txOi7oeT8AnD/Ogds0faue1GCrBBjaWJ3yLoPhlRTYv6bu5ZorMJFvdoih6J0O9sXbA gh6g== X-Gm-Message-State: AFqh2kqtt3FaHJW2aFBQ4NZWwgFkMlfxRJSJzIP6ACo2CtOQSIe8m8Fi qE4nztkbNi517wSDEmzgvY7kbA== X-Google-Smtp-Source: AMrXdXuHqFQNAZK7ZYvNrNDQyGG55zccD7RMvhGagAGAXDXY1I6p5WsQH0PLKbrSubkOpF2lxVjhow== X-Received: by 2002:aa7:d719:0:b0:474:a35b:6c17 with SMTP id t25-20020aa7d719000000b00474a35b6c17mr39821634edq.4.1672757724841; Tue, 03 Jan 2023 06:55:24 -0800 (PST) Received: from localhost.localdomain (mob-5-91-46-2.net.vodafone.it. [5.91.46.2]) by smtp.gmail.com with ESMTPSA id g17-20020a17090604d100b007c16fdc93ddsm14122595eja.81.2023.01.03.06.55.23 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Tue, 03 Jan 2023 06:55:24 -0800 (PST) From: Paolo Valente To: Jens Axboe Cc: linux-block@vger.kernel.org, linux-kernel@vger.kernel.org, arie.vanderhoeven@seagate.com, rory.c.chen@seagate.com, glen.valante@linaro.org, damien.lemoal@opensource.wdc.com, Paolo Valente , Gabriele Felici , Gianmarco Lusvardi , Giulio Barabino , Emiliano Maccaferri Subject: [PATCH V13 REBASED 4/8] block, bfq: turn bfqq_data into an array in bfq_io_cq Date: Tue, 3 Jan 2023 15:54:59 +0100 Message-Id: <20230103145503.71712-5-paolo.valente@linaro.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20230103145503.71712-1-paolo.valente@linaro.org> References: <20230103145503.71712-1-paolo.valente@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" When a bfq_queue Q is merged with another queue, several pieces of information are saved about Q. These pieces are stored in the bfqq_data field in the bfq_io_cq data structure of the process associated with Q. Yet, with a multi-actuator drive, a process may get associated with multiple bfq_queues: one queue for each of the N actuators. Each of these queues may undergo a merge. So, the bfq_io_cq data structure must be able to accommodate the above information for N queues. This commit solves this problem by turning the bfqq_data scalar field into an array of N elements (and by changing code so as to handle this array). This solution is written under the assumption that bfq_queues associated with different actuators cannot be cross-merged. This assumption holds naturally with basic queue merging: the latter is triggered by spatial locality, and sectors for different actuators are not close to each other (apart from the corner case of the last sectors served by a given actuator and the first sectors served by the next actuator). As for stable cross-merging, the assumption here is that it is disabled. Reviewed-by: Damien Le Moal Signed-off-by: Gabriele Felici Signed-off-by: Gianmarco Lusvardi Signed-off-by: Giulio Barabino Signed-off-by: Emiliano Maccaferri Signed-off-by: Paolo Valente --- block/bfq-iosched.c | 100 ++++++++++++++++++++++++++------------------ block/bfq-iosched.h | 12 ++++-- 2 files changed, 67 insertions(+), 45 deletions(-) diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index cefd16b9cbd1..9675a32af0f4 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -408,7 +408,7 @@ void bic_set_bfqq(struct bfq_io_cq *bic, * we cancel the stable merge if * bic->stable_merge_bfqq =3D=3D bfqq. */ - struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data[actuator_idx]; =20 /* Clear bic pointer if bfqq is detached from this bic */ if (old_bfqq && old_bfqq->bic =3D=3D bic) @@ -1135,9 +1135,10 @@ static void bfq_bfqq_resume_state(struct bfq_queue *bfqq, struct bfq_data *bfqd, struct bfq_io_cq *bic, bool bfq_already_existing) { - struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; unsigned int old_wr_coeff =3D 1; bool busy =3D bfq_already_existing && bfq_bfqq_busy(bfqq); + unsigned int a_idx =3D bfqq->actuator_idx; + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data[a_idx]; =20 if (bfqq_data->saved_has_short_ttime) bfq_mark_bfqq_has_short_ttime(bfqq); @@ -1818,7 +1819,9 @@ static void bfq_bfqq_handle_idle_busy_switch(struct b= fq_data *bfqd, arrived_in_time =3D ktime_get_ns() <=3D bfqq->ttime.last_end_request + bfqd->bfq_slice_idle * 3; - + unsigned int act_idx =3D bfq_actuator_index(bfqd, rq->bio); + bool bfqq_non_merged_or_stably_merged =3D + bfqq->bic || RQ_BIC(rq)->bfqq_data[act_idx].stably_merged; =20 /* * bfqq deserves to be weight-raised if: @@ -1852,9 +1855,8 @@ static void bfq_bfqq_handle_idle_busy_switch(struct b= fq_data *bfqd, */ wr_or_deserves_wr =3D bfqd->low_latency && (bfqq->wr_coeff > 1 || - (bfq_bfqq_sync(bfqq) && - (bfqq->bic || RQ_BIC(rq)->bfqq_data.stably_merged) && - (*interactive || soft_rt))); + (bfq_bfqq_sync(bfqq) && bfqq_non_merged_or_stably_merged && + (*interactive || soft_rt))); =20 /* * Using the last flag, update budget and check whether bfqq @@ -2828,6 +2830,35 @@ static bool bfq_may_be_close_cooperator(struct bfq_q= ueue *bfqq, static bool idling_boosts_thr_without_issues(struct bfq_data *bfqd, struct bfq_queue *bfqq); =20 +static struct bfq_queue * +bfq_setup_stable_merge(struct bfq_data *bfqd, struct bfq_queue *bfqq, + struct bfq_queue *stable_merge_bfqq, + struct bfq_iocq_bfqq_data *bfqq_data) +{ + int proc_ref =3D min(bfqq_process_refs(bfqq), + bfqq_process_refs(stable_merge_bfqq)); + struct bfq_queue *new_bfqq; + + if (idling_boosts_thr_without_issues(bfqd, bfqq) || + proc_ref =3D=3D 0) + return NULL; + + /* next function will take at least one ref */ + new_bfqq =3D bfq_setup_merge(bfqq, stable_merge_bfqq); + + if (new_bfqq) { + bfqq_data->stably_merged =3D true; + if (new_bfqq->bic) { + unsigned int new_a_idx =3D new_bfqq->actuator_idx; + struct bfq_iocq_bfqq_data *new_bfqq_data =3D + &new_bfqq->bic->bfqq_data[new_a_idx]; + + new_bfqq_data->stably_merged =3D true; + } + } + return new_bfqq; +} + /* * Attempt to schedule a merge of bfqq with the currently in-service * queue or with a close queue among the scheduled queues. Return @@ -2853,7 +2884,8 @@ bfq_setup_cooperator(struct bfq_data *bfqd, struct bf= q_queue *bfqq, void *io_struct, bool request, struct bfq_io_cq *bic) { struct bfq_queue *in_service_bfqq, *new_bfqq; - struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; + unsigned int a_idx =3D bfqq->actuator_idx; + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data[a_idx]; =20 /* if a merge has already been setup, then proceed with that first */ if (bfqq->new_bfqq) @@ -2883,29 +2915,15 @@ bfq_setup_cooperator(struct bfq_data *bfqd, struct = bfq_queue *bfqq, msecs_to_jiffies(bfq_late_stable_merging))) { struct bfq_queue *stable_merge_bfqq =3D bfqq_data->stable_merge_bfqq; - int proc_ref =3D min(bfqq_process_refs(bfqq), - bfqq_process_refs(stable_merge_bfqq)); =20 /* deschedule stable merge, because done or aborted here */ bfq_put_stable_ref(stable_merge_bfqq); =20 bfqq_data->stable_merge_bfqq =3D NULL; =20 - if (!idling_boosts_thr_without_issues(bfqd, bfqq) && - proc_ref > 0) { - /* next function will take at least one ref */ - struct bfq_queue *new_bfqq =3D - bfq_setup_merge(bfqq, stable_merge_bfqq); - - if (new_bfqq) { - bfqq_data->stably_merged =3D true; - if (new_bfqq->bic) - new_bfqq->bic->bfqq_data.stably_merged =3D - true; - } - return new_bfqq; - } else - return NULL; + return bfq_setup_stable_merge(bfqd, bfqq, + stable_merge_bfqq, + bfqq_data); } } =20 @@ -3000,7 +3018,8 @@ bfq_setup_cooperator(struct bfq_data *bfqd, struct bf= q_queue *bfqq, static void bfq_bfqq_save_state(struct bfq_queue *bfqq) { struct bfq_io_cq *bic =3D bfqq->bic; - struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; + unsigned int a_idx =3D bfqq->actuator_idx; + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data[a_idx]; =20 /* * If !bfqq->bic, the queue is already shared or its requests @@ -3011,7 +3030,7 @@ static void bfq_bfqq_save_state(struct bfq_queue *bfq= q) return; =20 bfqq_data->saved_last_serv_time_ns =3D bfqq->last_serv_time_ns; - bfqq_data->saved_inject_limit =3D bfqq->inject_limit; + bfqq_data->saved_inject_limit =3D bfqq->inject_limit; bfqq_data->saved_decrease_time_jif =3D bfqq->decrease_time_jif; =20 bfqq_data->saved_weight =3D bfqq->entity.orig_weight; @@ -5377,7 +5396,7 @@ static void bfq_exit_icq(struct io_cq *icq) * therefore on its unused per-actuator fields being NULL. */ unsigned int num_actuators =3D BFQ_MAX_ACTUATORS; - struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; + struct bfq_iocq_bfqq_data *bfqq_data =3D bic->bfqq_data; =20 /* * bfqd is NULL if scheduler already exited, and in that case @@ -5388,10 +5407,10 @@ static void bfq_exit_icq(struct io_cq *icq) num_actuators =3D bfqd->num_actuators; } =20 - if (bfqq_data->stable_merge_bfqq) - bfq_put_stable_ref(bfqq_data->stable_merge_bfqq); - for (act_idx =3D 0; act_idx < num_actuators; act_idx++) { + if (bfqq_data[act_idx].stable_merge_bfqq) + bfq_put_stable_ref(bfqq_data[act_idx].stable_merge_bfqq); + bfq_exit_icq_bfqq(bic, true, act_idx); bfq_exit_icq_bfqq(bic, false, act_idx); } @@ -5578,16 +5597,16 @@ bfq_do_early_stable_merge(struct bfq_data *bfqd, st= ruct bfq_queue *bfqq, struct bfq_io_cq *bic, struct bfq_queue *last_bfqq_created) { + unsigned int a_idx =3D last_bfqq_created->actuator_idx; struct bfq_queue *new_bfqq =3D bfq_setup_merge(bfqq, last_bfqq_created); - struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; =20 if (!new_bfqq) return bfqq; =20 if (new_bfqq->bic) - new_bfqq->bic->bfqq_data.stably_merged =3D true; - bfqq_data->stably_merged =3D true; + new_bfqq->bic->bfqq_data[a_idx].stably_merged =3D true; + bic->bfqq_data[a_idx].stably_merged =3D true; =20 /* * Reusing merge functions. This implies that @@ -5656,7 +5675,6 @@ static struct bfq_queue *bfq_do_or_sched_stable_merge= (struct bfq_data *bfqd, &bfqd->last_bfqq_created; =20 struct bfq_queue *last_bfqq_created =3D *source_bfqq; - struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; =20 /* * If last_bfqq_created has not been set yet, then init it. If @@ -5718,7 +5736,8 @@ static struct bfq_queue *bfq_do_or_sched_stable_merge= (struct bfq_data *bfqd, /* * Record the bfqq to merge to. */ - bfqq_data->stable_merge_bfqq =3D last_bfqq_created; + bic->bfqq_data[last_bfqq_created->actuator_idx].stable_merge_bfqq =3D + last_bfqq_created; } } =20 @@ -6639,7 +6658,7 @@ static struct bfq_queue *bfq_get_bfqq_handle_split(st= ruct bfq_data *bfqd, { unsigned int act_idx =3D bfq_actuator_index(bfqd, bio); struct bfq_queue *bfqq =3D bic_to_bfqq(bic, is_sync, act_idx); - struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data; + struct bfq_iocq_bfqq_data *bfqq_data =3D &bic->bfqq_data[act_idx]; =20 if (likely(bfqq && bfqq !=3D &bfqd->oom_bfqq)) return bfqq; @@ -6747,7 +6766,7 @@ static struct bfq_queue *bfq_init_rq(struct request *= rq) struct bfq_queue *bfqq; bool new_queue =3D false; bool bfqq_already_existing =3D false, split =3D false; - struct bfq_iocq_bfqq_data *bfqq_data; + unsigned int a_idx =3D bfq_actuator_index(bfqd, bio); =20 if (unlikely(!rq->elv.icq)) return NULL; @@ -6771,17 +6790,16 @@ static struct bfq_queue *bfq_init_rq(struct request= *rq) bfqq =3D bfq_get_bfqq_handle_split(bfqd, bic, bio, false, is_sync, &new_queue); =20 - bfqq_data =3D &bic->bfqq_data; - if (likely(!new_queue)) { /* If the queue was seeky for too long, break it apart. */ if (bfq_bfqq_coop(bfqq) && bfq_bfqq_split_coop(bfqq) && - !bfqq_data->stably_merged) { + !bic->bfqq_data[a_idx].stably_merged) { struct bfq_queue *old_bfqq =3D bfqq; =20 /* Update bic before losing reference to bfqq */ if (bfq_bfqq_in_large_burst(bfqq)) - bfqq_data->saved_in_large_burst =3D true; + bic->bfqq_data[a_idx].saved_in_large_burst =3D + true; =20 bfqq =3D bfq_split_bfqq(bic, bfqq); split =3D true; diff --git a/block/bfq-iosched.h b/block/bfq-iosched.h index 8e0215b8cc30..5a6f888930ad 100644 --- a/block/bfq-iosched.h +++ b/block/bfq-iosched.h @@ -418,7 +418,7 @@ struct bfq_queue { struct bfq_iocq_bfqq_data { /* * Snapshot of the has_short_time flag before merging; taken - * to remember its value while the queue is merged, so as to + * to remember its values while the queue is merged, so as to * be able to restore it in case of split. */ bool saved_has_short_ttime; @@ -432,7 +432,7 @@ struct bfq_iocq_bfqq_data { u64 saved_tot_idle_time; =20 /* - * Same purpose as the previous fields for the value of the + * Same purpose as the previous fields for the values of the * field keeping the queue's belonging to a large burst */ bool saved_in_large_burst; @@ -495,8 +495,12 @@ struct bfq_io_cq { uint64_t blkcg_serial_nr; /* the current blkcg serial */ #endif =20 - /* persistent data for associated synchronous process queue */ - struct bfq_iocq_bfqq_data bfqq_data; + /* + * Persistent data for associated synchronous process queues + * (one queue per actuator, see field bfqq above). In + * particular, each of these queues may undergo a merge. + */ + struct bfq_iocq_bfqq_data bfqq_data[BFQ_MAX_ACTUATORS]; =20 unsigned int requests; /* Number of requests this process has in flight */ }; --=20 2.20.1 From nobody Tue Sep 16 21:33:18 2025 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 4A9E9C46467 for ; Tue, 3 Jan 2023 14:56:00 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S233319AbjACOz5 (ORCPT ); Tue, 3 Jan 2023 09:55:57 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:60788 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237944AbjACOzc (ORCPT ); Tue, 3 Jan 2023 09:55:32 -0500 Received: from mail-ej1-x632.google.com (mail-ej1-x632.google.com [IPv6:2a00:1450:4864:20::632]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 46D86FD19 for ; Tue, 3 Jan 2023 06:55:31 -0800 (PST) Received: by mail-ej1-x632.google.com with SMTP id qk9so74210019ejc.3 for ; Tue, 03 Jan 2023 06:55:31 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=Ia/o1TRvqYroHXBLDuv3zNcB1AlF7EjIkmp8+qRWGx4=; b=Q2mLTn6+vYMncdNXqI5oGBQzLnF1xJQkxcfDAugyWiA49wdfTot4RxgErceq75Q4op RY0qfQh4a4OWsf6t2/sgKDEZ1yWUex5qEXKFEJBS0GLcjI0Wq4OOcfoOSxxy5vUiIwCf l9adK9xHEsxCZpBDrO9nCT7sdK3s6QcK76IlF8mbjPHmx7C605YXBCGhomed1nM8E/l3 Vb5Qb67AMKKSfeORtnKqCbibFdVP2wqgt9kEh3H3Kuo/rDYHcYckUzvKe5ASn5DZRPrs IxSRwDZYP1AmYtUM5OrP0kDApvIJpZG5+z7jxq8l6PCRU2bhw/2pTfGua/T2RoXmXCAb 1D7A== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=Ia/o1TRvqYroHXBLDuv3zNcB1AlF7EjIkmp8+qRWGx4=; b=AkNme/p4RBxOTvhw9S97yqC7FWCc1hFb1QKYwal99usWojjxPnvXgz2z3ij0XrKgZD +5T+NqjDdiZ8ZY83wDLItr6v/b8cbRdE+sTAu/gVB1ev1Urrbxg2lIMncGLLCStUrB0K RTAeYQIQ1CLPBrCe18ynbUeP/jLcqYQb0Gy15yG+l9nL8/AbjeG2XWipatDaKwR6+LBV vcKKLuIKa0visVabD8ItHs1LVi1yagkoGOyxn/7Wb1nUy9umKL/SJ7B8ggEwG1kL6zGb UJOcqinll2Od4CipRMnU/OVIyKvnjjAosGtrTS1gxxgYKsdxoR3vax2PLqDGwu4j6Wv3 9lAA== X-Gm-Message-State: AFqh2ko4uVIMBCGlQzFGEBdDbdzCA/ofkQckURBbQyJ6zOylaIz5DLhp TUvbnlsHSmHmRaCBUyw+Z7ZSOA== X-Google-Smtp-Source: AMrXdXvHwQTTBT6srp/7zJbFaMc2hyL9cAdnyfpOaZTMTIx1yJNZ4L+e6sIYRZrpo5De+rVq+P2NpA== X-Received: by 2002:a17:906:1ccc:b0:7c0:d88b:4323 with SMTP id i12-20020a1709061ccc00b007c0d88b4323mr33040365ejh.61.1672757729794; Tue, 03 Jan 2023 06:55:29 -0800 (PST) Received: from localhost.localdomain (mob-5-91-46-2.net.vodafone.it. [5.91.46.2]) by smtp.gmail.com with ESMTPSA id g17-20020a17090604d100b007c16fdc93ddsm14122595eja.81.2023.01.03.06.55.25 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Tue, 03 Jan 2023 06:55:29 -0800 (PST) From: Paolo Valente To: Jens Axboe Cc: linux-block@vger.kernel.org, linux-kernel@vger.kernel.org, arie.vanderhoeven@seagate.com, rory.c.chen@seagate.com, glen.valante@linaro.org, damien.lemoal@opensource.wdc.com, Davide Zini , Paolo Valente Subject: [PATCH V13 REBASED 5/8] block, bfq: split also async bfq_queues on a per-actuator basis Date: Tue, 3 Jan 2023 15:55:00 +0100 Message-Id: <20230103145503.71712-6-paolo.valente@linaro.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20230103145503.71712-1-paolo.valente@linaro.org> References: <20230103145503.71712-1-paolo.valente@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" From: Davide Zini Similarly to sync bfq_queues, also async bfq_queues need to be split on a per-actuator basis. Reviewed-by: Damien Le Moal Signed-off-by: Paolo Valente Signed-off-by: Davide Zini --- block/bfq-iosched.c | 41 +++++++++++++++++++++++------------------ block/bfq-iosched.h | 8 ++++---- 2 files changed, 27 insertions(+), 22 deletions(-) diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index 9675a32af0f4..0859981726ac 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -2620,14 +2620,16 @@ static void bfq_bfqq_end_wr(struct bfq_queue *bfqq) void bfq_end_wr_async_queues(struct bfq_data *bfqd, struct bfq_group *bfqg) { - int i, j; - - for (i =3D 0; i < 2; i++) - for (j =3D 0; j < IOPRIO_NR_LEVELS; j++) - if (bfqg->async_bfqq[i][j]) - bfq_bfqq_end_wr(bfqg->async_bfqq[i][j]); - if (bfqg->async_idle_bfqq) - bfq_bfqq_end_wr(bfqg->async_idle_bfqq); + int i, j, k; + + for (k =3D 0; k < bfqd->num_actuators; k++) { + for (i =3D 0; i < 2; i++) + for (j =3D 0; j < IOPRIO_NR_LEVELS; j++) + if (bfqg->async_bfqq[i][j][k]) + bfq_bfqq_end_wr(bfqg->async_bfqq[i][j][k]); + if (bfqg->async_idle_bfqq[k]) + bfq_bfqq_end_wr(bfqg->async_idle_bfqq[k]); + } } =20 static void bfq_end_wr(struct bfq_data *bfqd) @@ -5575,18 +5577,18 @@ static void bfq_init_bfqq(struct bfq_data *bfqd, st= ruct bfq_queue *bfqq, =20 static struct bfq_queue **bfq_async_queue_prio(struct bfq_data *bfqd, struct bfq_group *bfqg, - int ioprio_class, int ioprio) + int ioprio_class, int ioprio, int act_idx) { switch (ioprio_class) { case IOPRIO_CLASS_RT: - return &bfqg->async_bfqq[0][ioprio]; + return &bfqg->async_bfqq[0][ioprio][act_idx]; case IOPRIO_CLASS_NONE: ioprio =3D IOPRIO_BE_NORM; fallthrough; case IOPRIO_CLASS_BE: - return &bfqg->async_bfqq[1][ioprio]; + return &bfqg->async_bfqq[1][ioprio][act_idx]; case IOPRIO_CLASS_IDLE: - return &bfqg->async_idle_bfqq; + return &bfqg->async_idle_bfqq[act_idx]; default: return NULL; } @@ -5759,7 +5761,8 @@ static struct bfq_queue *bfq_get_queue(struct bfq_dat= a *bfqd, bfqg =3D bfq_bio_bfqg(bfqd, bio); if (!is_sync) { async_bfqq =3D bfq_async_queue_prio(bfqd, bfqg, ioprio_class, - ioprio); + ioprio, + bfq_actuator_index(bfqd, bio)); bfqq =3D *async_bfqq; if (bfqq) goto out; @@ -6982,13 +6985,15 @@ static void __bfq_put_async_bfqq(struct bfq_data *b= fqd, */ void bfq_put_async_queues(struct bfq_data *bfqd, struct bfq_group *bfqg) { - int i, j; + int i, j, k; =20 - for (i =3D 0; i < 2; i++) - for (j =3D 0; j < IOPRIO_NR_LEVELS; j++) - __bfq_put_async_bfqq(bfqd, &bfqg->async_bfqq[i][j]); + for (k =3D 0; k < bfqd->num_actuators; k++) { + for (i =3D 0; i < 2; i++) + for (j =3D 0; j < IOPRIO_NR_LEVELS; j++) + __bfq_put_async_bfqq(bfqd, &bfqg->async_bfqq[i][j][k]); =20 - __bfq_put_async_bfqq(bfqd, &bfqg->async_idle_bfqq); + __bfq_put_async_bfqq(bfqd, &bfqg->async_idle_bfqq[k]); + } } =20 /* diff --git a/block/bfq-iosched.h b/block/bfq-iosched.h index 5a6f888930ad..de2b2af643e5 100644 --- a/block/bfq-iosched.h +++ b/block/bfq-iosched.h @@ -980,8 +980,8 @@ struct bfq_group { =20 struct bfq_data *bfqd; =20 - struct bfq_queue *async_bfqq[2][IOPRIO_NR_LEVELS]; - struct bfq_queue *async_idle_bfqq; + struct bfq_queue *async_bfqq[2][IOPRIO_NR_LEVELS][BFQ_MAX_ACTUATORS]; + struct bfq_queue *async_idle_bfqq[BFQ_MAX_ACTUATORS]; =20 struct bfq_entity *my_entity; =20 @@ -998,8 +998,8 @@ struct bfq_group { struct bfq_entity entity; struct bfq_sched_data sched_data; =20 - struct bfq_queue *async_bfqq[2][IOPRIO_NR_LEVELS]; - struct bfq_queue *async_idle_bfqq; + struct bfq_queue *async_bfqq[2][IOPRIO_NR_LEVELS][BFQ_MAX_ACTUATORS]; + struct bfq_queue *async_idle_bfqq[BFQ_MAX_ACTUATORS]; =20 struct rb_root rq_pos_tree; }; --=20 2.20.1 From nobody Tue Sep 16 21:33:18 2025 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 8A35CC46467 for ; Tue, 3 Jan 2023 14:56:16 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S238106AbjACO4O (ORCPT ); Tue, 3 Jan 2023 09:56:14 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:60968 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S233372AbjACOzo (ORCPT ); Tue, 3 Jan 2023 09:55:44 -0500 Received: from mail-ej1-x629.google.com (mail-ej1-x629.google.com [IPv6:2a00:1450:4864:20::629]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 3F066FD0C for ; Tue, 3 Jan 2023 06:55:33 -0800 (PST) Received: by mail-ej1-x629.google.com with SMTP id tz12so74283903ejc.9 for ; Tue, 03 Jan 2023 06:55:33 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=fsWPGWHz4PsB9F2bSORsl2cpLwLReathmqV+xA12OGo=; b=nKMNURzhoK5XU7f8rSpM3To47grIJM+ZHDwjyFQMf5NMCYcosDi/AdcodXD0Zkm+gM yjpr1CjTaFvFRNyLLsrtZbikyS7D7lFKmlqXDfr+FhfgHU0nUrZD1LOv9lxTrkkzZ+ws sJ4vNN7ILBu4yLH32zNNAGqslw8WIVD2ZlE4PRAV5S8E7ezYL5V/0DuZ9/ysH334iISt aaQE+Z+3d2N1oEIxhi8CSejLYhV8oZfU9acNgzsYXwDOekx1fZmLf6Yhj+f6aW0EPfVw idN429IAiPtTduoEKQI8OMUWkODhbRkw2V4Fi43lZLDU1HnpRNG8oO+VfNkEcx0B1NFq O0Mg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=fsWPGWHz4PsB9F2bSORsl2cpLwLReathmqV+xA12OGo=; b=oFjKFwH4Ei4AJiNMIFK0k7qY5dEufsIRaCD2l5sJx56OjLl0QCuUH3TcpcdgzuNo4F Su56/aQHLfeK1kwHmHv8ZRfMTMhrdKuq3cEQzSjFyqDaruAddJY+4H8Qot0zhTdOh16p iY4LR+tq8z0Zxv+/z7koKxXPDH9aQnENzLSaL/ZnyccBMITrTdMvmfr1Vx8yvCfGettd OSnfSjJGhGRO260r/XKAUA9Jaw2miyiWo6YTcVlJ78G0YYOaS10AH1a6zOPY3fXYk9t6 yxw4A27iSmzJ0GCIhNTdcZqcuYQRT0WKJyeMRyK6SghO3DUCO8D5SeVVZuoPMqDRCWfW XEQg== X-Gm-Message-State: AFqh2krGuFsnl64cLLpv7uqyaBSdLDnynMSYhcloFFsU73+Y0y2M7ZfY f0ZZJwQ5KnmSTR73AvQ9zpxsZA== X-Google-Smtp-Source: AMrXdXs02M/o1s6jHZ/t3xeo0oz6Mq1A59MbGXMJlqgPfxejObu5lu22+cdXtYO1Iw0zO3Sr8Sgnwg== X-Received: by 2002:a17:906:7fc3:b0:7c1:10b8:e6a4 with SMTP id r3-20020a1709067fc300b007c110b8e6a4mr35703060ejs.19.1672757731786; Tue, 03 Jan 2023 06:55:31 -0800 (PST) Received: from localhost.localdomain (mob-5-91-46-2.net.vodafone.it. [5.91.46.2]) by smtp.gmail.com with ESMTPSA id g17-20020a17090604d100b007c16fdc93ddsm14122595eja.81.2023.01.03.06.55.29 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Tue, 03 Jan 2023 06:55:31 -0800 (PST) From: Paolo Valente To: Jens Axboe Cc: linux-block@vger.kernel.org, linux-kernel@vger.kernel.org, arie.vanderhoeven@seagate.com, rory.c.chen@seagate.com, glen.valante@linaro.org, damien.lemoal@opensource.wdc.com, Federico Gavioli , Paolo Valente Subject: [PATCH V13 REBASED 6/8] block, bfq: retrieve independent access ranges from request queue Date: Tue, 3 Jan 2023 15:55:01 +0100 Message-Id: <20230103145503.71712-7-paolo.valente@linaro.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20230103145503.71712-1-paolo.valente@linaro.org> References: <20230103145503.71712-1-paolo.valente@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" From: Federico Gavioli This patch implements the code to gather the content of the independent_access_ranges structure from the request_queue and copy it into the queue's bfq_data. This copy is done at queue initialization. We copy the access ranges into the bfq_data to avoid taking the queue lock each time we access the ranges. This implementation, however, puts a limit to the maximum independent ranges supported by the scheduler. Such a limit is equal to the constant BFQ_MAX_ACTUATORS. This limit was placed to avoid the allocation of dynamic memory. Reviewed-by: Damien Le Moal Co-developed-by: Rory Chen Signed-off-by: Rory Chen Signed-off-by: Federico Gavioli Signed-off-by: Paolo Valente --- block/bfq-iosched.c | 59 +++++++++++++++++++++++++++++++++++++++------ block/bfq-iosched.h | 8 +++++- 2 files changed, 58 insertions(+), 9 deletions(-) diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index 0859981726ac..56486f24b4c5 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -1793,10 +1793,25 @@ static bool bfq_bfqq_higher_class_or_weight(struct = bfq_queue *bfqq, */ static unsigned int bfq_actuator_index(struct bfq_data *bfqd, struct bio *= bio) { - /* - * Multi-actuator support not complete yet, so always return 0 - * for the moment (to keep incomplete mechanisms off). - */ + unsigned int i; + sector_t end; + + /* no search needed if one or zero ranges present */ + if (bfqd->num_actuators =3D=3D 1) + return 0; + + /* bio_end_sector(bio) gives the sector after the last one */ + end =3D bio_end_sector(bio) - 1; + + for (i =3D 0; i < bfqd->num_actuators; i++) { + if (end >=3D bfqd->sector[i] && + end < bfqd->sector[i] + bfqd->nr_sectors[i]) + return i; + } + + WARN_ONCE(true, + "bfq_actuator_index: bio sector out of ranges: end=3D%llu\n", + end); return 0; } =20 @@ -7105,6 +7120,8 @@ static int bfq_init_queue(struct request_queue *q, st= ruct elevator_type *e) { struct bfq_data *bfqd; struct elevator_queue *eq; + unsigned int i; + struct blk_independent_access_ranges *ia_ranges =3D q->disk->ia_ranges; =20 eq =3D elevator_alloc(q, e); if (!eq) @@ -7147,12 +7164,38 @@ static int bfq_init_queue(struct request_queue *q, = struct elevator_type *e) =20 bfqd->queue =3D q; =20 + bfqd->num_actuators =3D 1; /* - * Multi-actuator support not complete yet, unconditionally - * set to only one actuator for the moment (to keep incomplete - * mechanisms off). + * If the disk supports multiple actuators, copy independent + * access ranges from the request queue structure. */ - bfqd->num_actuators =3D 1; + spin_lock_irq(&q->queue_lock); + if (ia_ranges) { + /* + * Check if the disk ia_ranges size exceeds the current bfq + * actuator limit. + */ + if (ia_ranges->nr_ia_ranges > BFQ_MAX_ACTUATORS) { + pr_crit("nr_ia_ranges higher than act limit: iars=3D%d, max=3D%d.\n", + ia_ranges->nr_ia_ranges, BFQ_MAX_ACTUATORS); + pr_crit("Falling back to single actuator mode.\n"); + } else { + bfqd->num_actuators =3D ia_ranges->nr_ia_ranges; + + for (i =3D 0; i < bfqd->num_actuators; i++) { + bfqd->sector[i] =3D ia_ranges->ia_range[i].sector; + bfqd->nr_sectors[i] =3D + ia_ranges->ia_range[i].nr_sectors; + } + } + } + + /* Otherwise use single-actuator dev info */ + if (bfqd->num_actuators =3D=3D 1) { + bfqd->sector[0] =3D 0; + bfqd->nr_sectors[0] =3D get_capacity(q->disk); + } + spin_unlock_irq(&q->queue_lock); =20 INIT_LIST_HEAD(&bfqd->dispatch); =20 diff --git a/block/bfq-iosched.h b/block/bfq-iosched.h index de2b2af643e5..830dda1f9322 100644 --- a/block/bfq-iosched.h +++ b/block/bfq-iosched.h @@ -814,7 +814,13 @@ struct bfq_data { * case of single-actuator drives. */ unsigned int num_actuators; - + /* + * Disk independent access ranges for each actuator + * in this device. + */ + sector_t sector[BFQ_MAX_ACTUATORS]; + sector_t nr_sectors[BFQ_MAX_ACTUATORS]; + struct blk_independent_access_range ia_ranges[BFQ_MAX_ACTUATORS]; }; =20 enum bfqq_state_flags { --=20 2.20.1 From nobody Tue Sep 16 21:33:18 2025 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 B8C51C3DA7D for ; Tue, 3 Jan 2023 14:56:13 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S238099AbjACO4L (ORCPT ); Tue, 3 Jan 2023 09:56:11 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:60636 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S238000AbjACOzp (ORCPT ); Tue, 3 Jan 2023 09:55:45 -0500 Received: from mail-ej1-x636.google.com (mail-ej1-x636.google.com [IPv6:2a00:1450:4864:20::636]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id BA0E41208A for ; Tue, 3 Jan 2023 06:55:33 -0800 (PST) Received: by mail-ej1-x636.google.com with SMTP id fy8so10337634ejc.13 for ; Tue, 03 Jan 2023 06:55:33 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=GXXK2qnu5qVvXVUzy+cOnV+H+KufbHHn0u7i4BD209g=; b=i+X0wiT7210gip+CLvAq6qmtBCpGBb66d06z9pRHvtO4FY6W/cWDN6KOFSF//XV64G VoHBRXE9oQKKFBQ5jLRQ4ix1173gEwGa01Sz2IsNgTrCxR/XlBXyDMJx9mWuydjrXvmZ /g4js18lS+pQPk094UuDqVdMvZTwbChzklZZya4Q+1CVdBuVgaffwvQPDqC9u/KPpWIV wXfzTRf8u+H9L9zvxNFMhsAYJraOrCmd7W1v/4IzbnemN88MQwtXVeV7xlfZqJZpocc4 8r+oQrZ0Lv0ODgqbwNy4uc4WUo5Mpo7jJQx2ZeATE04hIgfy2ljEV4vYosxtPYW+RFt9 vxGw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=GXXK2qnu5qVvXVUzy+cOnV+H+KufbHHn0u7i4BD209g=; b=WevRk/XwIftF3NXOSc//bR6OuKV5IsV16dOp7Agn5y1l+yHVm+oYQuH7Odp48m8ymn gWfrveRutcs4fny4ej2X/jOIUEzqn3fKkR1JPNfMa+cowTTf2jvCaftqWTLufA+qK+Sa gB8yB5HDr95YsJf/+5MAbYW3osnNrbGCcqElFa11vmQJY/gPFfF5WEtYCkjyKNmIe82I cx+NwAFedEJVjGZ/WUTZK3rftSGRc2RG49iWhmxNvhRULg6w4AnqhSwwUdlUHDSeCIDE 4rSVMJJKs/sGK0RgJk6XSeF5nI0yvuhsKZ4U34Qdx8KFp8//ikibtdzcWxm9XbG2zad6 g2pA== X-Gm-Message-State: AFqh2kom6oiHRlcddwQQ23VYZ90DI2sa5alhk8d2YSDE9wnQzofSlH7q ymQSLBiXOZIJVsnb0EuXaLxlaQ== X-Google-Smtp-Source: AMrXdXsGo1CQJbc5pGrekNbPr2X7RkIwLWOLTpG0+cdR4ggHhBxHfIMVWquBMoZOuqkQF/flcqrUJQ== X-Received: by 2002:a17:906:5048:b0:7c0:b770:df94 with SMTP id e8-20020a170906504800b007c0b770df94mr38506880ejk.63.1672757733143; Tue, 03 Jan 2023 06:55:33 -0800 (PST) Received: from localhost.localdomain (mob-5-91-46-2.net.vodafone.it. [5.91.46.2]) by smtp.gmail.com with ESMTPSA id g17-20020a17090604d100b007c16fdc93ddsm14122595eja.81.2023.01.03.06.55.31 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Tue, 03 Jan 2023 06:55:32 -0800 (PST) From: Paolo Valente To: Jens Axboe Cc: linux-block@vger.kernel.org, linux-kernel@vger.kernel.org, arie.vanderhoeven@seagate.com, rory.c.chen@seagate.com, glen.valante@linaro.org, damien.lemoal@opensource.wdc.com, Davide Zini , Paolo Valente Subject: [PATCH V13 REBASED 7/8] block, bfq: inject I/O to underutilized actuators Date: Tue, 3 Jan 2023 15:55:02 +0100 Message-Id: <20230103145503.71712-8-paolo.valente@linaro.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20230103145503.71712-1-paolo.valente@linaro.org> References: <20230103145503.71712-1-paolo.valente@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" From: Davide Zini The main service scheme of BFQ for sync I/O is serving one sync bfq_queue at a time, for a while. In particular, BFQ enforces this scheme when it deems the latter necessary to boost throughput or to preserve service guarantees. Unfortunately, when BFQ enforces this policy, only one actuator at a time gets served for a while, because each bfq_queue contains I/O only for one actuator. The other actuators may remain underutilized. Actually, BFQ may serve (inject) extra I/O, taken from other bfq_queues, in parallel with that of the in-service queue. This injection mechanism may provide the ground for dealing also with the above actuator-underutilization problem. Yet BFQ does not take the actuator load into account when choosing which queue to pick extra I/O from. In addition, BFQ may happen to inject extra I/O only when the in-service queue is temporarily empty. In view of these facts, this commit extends the injection mechanism in such a way that the latter: (1) takes into account also the actuator load; (2) checks such a load on each dispatch, and injects I/O for an underutilized actuator, if there is one and there is I/O for it. To perform the check in (2), this commit introduces a load threshold, currently set to 4. A linear scan of each actuator is performed, until an actuator is found for which the following two conditions hold: the load of the actuator is below the threshold, and there is at least one non-in-service queue that contains I/O for that actuator. If such a pair (actuator, queue) is found, then the head request of that queue is returned for dispatch, instead of the head request of the in-service queue. We have set the threshold, empirically, to the minimum possible value for which an actuator is fully utilized, or close to be fully utilized. By doing so, injected I/O 'steals' as few drive-queue slots as possibile to the in-service queue. This reduces as much as possible the probability that the service of I/O from the in-service bfq_queue gets delayed because of slot exhaustion, i.e., because all the slots of the drive queue are filled with I/O injected from other queues (NCQ provides for 32 slots). This new mechanism also counters actuator underutilization in the case of asymmetric configurations of bfq_queues. Namely if there are few bfq_queues containing I/O for some actuators and many bfq_queues containing I/O for other actuators. Or if the bfq_queues containing I/O for some actuators have lower weights than the other bfq_queues. Reviewed-by: Damien Le Moal Signed-off-by: Paolo Valente Signed-off-by: Davide Zini --- block/bfq-cgroup.c | 2 +- block/bfq-iosched.c | 136 ++++++++++++++++++++++++++++++++------------ block/bfq-iosched.h | 39 ++++++++++++- block/bfq-wf2q.c | 2 +- 4 files changed, 139 insertions(+), 40 deletions(-) diff --git a/block/bfq-cgroup.c b/block/bfq-cgroup.c index bd7bf0f8121d..a6e8da5f5cfd 100644 --- a/block/bfq-cgroup.c +++ b/block/bfq-cgroup.c @@ -708,7 +708,7 @@ void bfq_bfqq_move(struct bfq_data *bfqd, struct bfq_qu= eue *bfqq, bfq_activate_bfqq(bfqd, bfqq); } =20 - if (!bfqd->in_service_queue && !bfqd->rq_in_driver) + if (!bfqd->in_service_queue && !bfqd->tot_rq_in_driver) bfq_schedule_dispatch(bfqd); /* release extra ref taken above, bfqq may happen to be freed now */ bfq_put_queue(bfqq); diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index 56486f24b4c5..d42a229b5a86 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -2259,9 +2259,9 @@ static void bfq_add_request(struct request *rq) * elapsed. */ if (bfqq =3D=3D bfqd->in_service_queue && - (bfqd->rq_in_driver =3D=3D 0 || + (bfqd->tot_rq_in_driver =3D=3D 0 || (bfqq->last_serv_time_ns > 0 && - bfqd->rqs_injected && bfqd->rq_in_driver > 0)) && + bfqd->rqs_injected && bfqd->tot_rq_in_driver > 0)) && time_is_before_eq_jiffies(bfqq->decrease_time_jif + msecs_to_jiffies(10))) { bfqd->last_empty_occupied_ns =3D ktime_get_ns(); @@ -2285,7 +2285,7 @@ static void bfq_add_request(struct request *rq) * will be set in case injection is performed * on bfqq before rq is completed). */ - if (bfqd->rq_in_driver =3D=3D 0) + if (bfqd->tot_rq_in_driver =3D=3D 0) bfqd->rqs_injected =3D false; } } @@ -2650,11 +2650,14 @@ void bfq_end_wr_async_queues(struct bfq_data *bfqd, static void bfq_end_wr(struct bfq_data *bfqd) { struct bfq_queue *bfqq; + int i; =20 spin_lock_irq(&bfqd->lock); =20 - list_for_each_entry(bfqq, &bfqd->active_list, bfqq_list) - bfq_bfqq_end_wr(bfqq); + for (i =3D 0; i < bfqd->num_actuators; i++) { + list_for_each_entry(bfqq, &bfqd->active_list[i], bfqq_list) + bfq_bfqq_end_wr(bfqq); + } list_for_each_entry(bfqq, &bfqd->idle_list, bfqq_list) bfq_bfqq_end_wr(bfqq); bfq_end_wr_async(bfqd); @@ -3611,13 +3614,13 @@ static void bfq_update_peak_rate(struct bfq_data *b= fqd, struct request *rq) * - start a new observation interval with this dispatch */ if (now_ns - bfqd->last_dispatch > 100*NSEC_PER_MSEC && - bfqd->rq_in_driver =3D=3D 0) + bfqd->tot_rq_in_driver =3D=3D 0) goto update_rate_and_reset; =20 /* Update sampling information */ bfqd->peak_rate_samples++; =20 - if ((bfqd->rq_in_driver > 0 || + if ((bfqd->tot_rq_in_driver > 0 || now_ns - bfqd->last_completion < BFQ_MIN_TT) && !BFQ_RQ_SEEKY(bfqd, bfqd->last_position, rq)) bfqd->sequential_samples++; @@ -3882,10 +3885,8 @@ static bool idling_needed_for_service_guarantees(str= uct bfq_data *bfqd, return false; =20 return (bfqq->wr_coeff > 1 && - (bfqd->wr_busy_queues < - tot_busy_queues || - bfqd->rq_in_driver >=3D - bfqq->dispatched + 4)) || + (bfqd->wr_busy_queues < tot_busy_queues || + bfqd->tot_rq_in_driver >=3D bfqq->dispatched + 4)) || bfq_asymmetric_scenario(bfqd, bfqq) || tot_busy_queues =3D=3D 1; } @@ -4656,6 +4657,8 @@ bfq_choose_bfqq_for_injection(struct bfq_data *bfqd) { struct bfq_queue *bfqq, *in_serv_bfqq =3D bfqd->in_service_queue; unsigned int limit =3D in_serv_bfqq->inject_limit; + int i; + /* * If * - bfqq is not weight-raised and therefore does not carry @@ -4687,7 +4690,7 @@ bfq_choose_bfqq_for_injection(struct bfq_data *bfqd) ) limit =3D 1; =20 - if (bfqd->rq_in_driver >=3D limit) + if (bfqd->tot_rq_in_driver >=3D limit) return NULL; =20 /* @@ -4702,11 +4705,12 @@ bfq_choose_bfqq_for_injection(struct bfq_data *bfqd) * (and re-added only if it gets new requests, but then it * is assigned again enough budget for its new backlog). */ - list_for_each_entry(bfqq, &bfqd->active_list, bfqq_list) - if (!RB_EMPTY_ROOT(&bfqq->sort_list) && - (in_serv_always_inject || bfqq->wr_coeff > 1) && - bfq_serv_to_charge(bfqq->next_rq, bfqq) <=3D - bfq_bfqq_budget_left(bfqq)) { + for (i =3D 0; i < bfqd->num_actuators; i++) { + list_for_each_entry(bfqq, &bfqd->active_list[i], bfqq_list) + if (!RB_EMPTY_ROOT(&bfqq->sort_list) && + (in_serv_always_inject || bfqq->wr_coeff > 1) && + bfq_serv_to_charge(bfqq->next_rq, bfqq) <=3D + bfq_bfqq_budget_left(bfqq)) { /* * Allow for only one large in-flight request * on non-rotational devices, for the @@ -4731,22 +4735,69 @@ bfq_choose_bfqq_for_injection(struct bfq_data *bfqd) else limit =3D in_serv_bfqq->inject_limit; =20 - if (bfqd->rq_in_driver < limit) { + if (bfqd->tot_rq_in_driver < limit) { bfqd->rqs_injected =3D true; return bfqq; } } + } + + return NULL; +} + +static struct bfq_queue * +bfq_find_active_bfqq_for_actuator(struct bfq_data *bfqd, int idx) +{ + struct bfq_queue *bfqq; + + if (bfqd->in_service_queue && + bfqd->in_service_queue->actuator_idx =3D=3D idx) + return bfqd->in_service_queue; + + list_for_each_entry(bfqq, &bfqd->active_list[idx], bfqq_list) { + if (!RB_EMPTY_ROOT(&bfqq->sort_list) && + bfq_serv_to_charge(bfqq->next_rq, bfqq) <=3D + bfq_bfqq_budget_left(bfqq)) { + return bfqq; + } + } =20 return NULL; } =20 +/* + * Perform a linear scan of each actuator, until an actuator is found + * for which the following two conditions hold: the load of the + * actuator is below the threshold (see comments on actuator_load_threshold + * for details), and there is a queue that contains I/O for that + * actuator. On success, return that queue. + */ +static struct bfq_queue * +bfq_find_bfqq_for_underused_actuator(struct bfq_data *bfqd) +{ + int i; + + for (i =3D 0 ; i < bfqd->num_actuators; i++) { + if (bfqd->rq_in_driver[i] < bfqd->actuator_load_threshold) { + struct bfq_queue *bfqq =3D + bfq_find_active_bfqq_for_actuator(bfqd, i); + + if (bfqq) + return bfqq; + } + } + + return NULL; +} + + /* * Select a queue for service. If we have a current queue in service, * check whether to continue servicing it, or retrieve and set a new one. */ static struct bfq_queue *bfq_select_queue(struct bfq_data *bfqd) { - struct bfq_queue *bfqq; + struct bfq_queue *bfqq, *inject_bfqq; struct request *next_rq; enum bfqq_expiration reason =3D BFQQE_BUDGET_TIMEOUT; =20 @@ -4768,6 +4819,15 @@ static struct bfq_queue *bfq_select_queue(struct bfq= _data *bfqd) goto expire; =20 check_queue: + /* + * If some actuator is underutilized, but the in-service + * queue does not contain I/O for that actuator, then try to + * inject I/O for that actuator. + */ + inject_bfqq =3D bfq_find_bfqq_for_underused_actuator(bfqd); + if (inject_bfqq && inject_bfqq !=3D bfqq) + return inject_bfqq; + /* * This loop is rarely executed more than once. Even when it * happens, it is much more convenient to re-execute this loop @@ -5123,11 +5183,11 @@ static struct request *__bfq_dispatch_request(struc= t blk_mq_hw_ctx *hctx) =20 /* * We exploit the bfq_finish_requeue_request hook to - * decrement rq_in_driver, but + * decrement tot_rq_in_driver, but * bfq_finish_requeue_request will not be invoked on * this request. So, to avoid unbalance, just start - * this request, without incrementing rq_in_driver. As - * a negative consequence, rq_in_driver is deceptively + * this request, without incrementing tot_rq_in_driver. As + * a negative consequence, tot_rq_in_driver is deceptively * lower than it should be while this request is in * service. This may cause bfq_schedule_dispatch to be * invoked uselessly. @@ -5136,7 +5196,7 @@ static struct request *__bfq_dispatch_request(struct = blk_mq_hw_ctx *hctx) * bfq_finish_requeue_request hook, if defined, is * probably invoked also on this request. So, by * exploiting this hook, we could 1) increment - * rq_in_driver here, and 2) decrement it in + * tot_rq_in_driver here, and 2) decrement it in * bfq_finish_requeue_request. Such a solution would * let the value of the counter be always accurate, * but it would entail using an extra interface @@ -5165,7 +5225,7 @@ static struct request *__bfq_dispatch_request(struct = blk_mq_hw_ctx *hctx) * Of course, serving one request at a time may cause loss of * throughput. */ - if (bfqd->strict_guarantees && bfqd->rq_in_driver > 0) + if (bfqd->strict_guarantees && bfqd->tot_rq_in_driver > 0) goto exit; =20 bfqq =3D bfq_select_queue(bfqd); @@ -5176,7 +5236,8 @@ static struct request *__bfq_dispatch_request(struct = blk_mq_hw_ctx *hctx) =20 if (rq) { inc_in_driver_start_rq: - bfqd->rq_in_driver++; + bfqd->rq_in_driver[bfqq->actuator_idx]++; + bfqd->tot_rq_in_driver++; start_rq: rq->rq_flags |=3D RQF_STARTED; } @@ -6243,7 +6304,7 @@ static void bfq_update_hw_tag(struct bfq_data *bfqd) struct bfq_queue *bfqq =3D bfqd->in_service_queue; =20 bfqd->max_rq_in_driver =3D max_t(int, bfqd->max_rq_in_driver, - bfqd->rq_in_driver); + bfqd->tot_rq_in_driver); =20 if (bfqd->hw_tag =3D=3D 1) return; @@ -6254,7 +6315,7 @@ static void bfq_update_hw_tag(struct bfq_data *bfqd) * sum is not exact, as it's not taking into account deactivated * requests. */ - if (bfqd->rq_in_driver + bfqd->queued <=3D BFQ_HW_QUEUE_THRESHOLD) + if (bfqd->tot_rq_in_driver + bfqd->queued <=3D BFQ_HW_QUEUE_THRESHOLD) return; =20 /* @@ -6265,7 +6326,7 @@ static void bfq_update_hw_tag(struct bfq_data *bfqd) if (bfqq && bfq_bfqq_has_short_ttime(bfqq) && bfqq->dispatched + bfqq->queued[0] + bfqq->queued[1] < BFQ_HW_QUEUE_THRESHOLD && - bfqd->rq_in_driver < BFQ_HW_QUEUE_THRESHOLD) + bfqd->tot_rq_in_driver < BFQ_HW_QUEUE_THRESHOLD) return; =20 if (bfqd->hw_tag_samples++ < BFQ_HW_QUEUE_SAMPLES) @@ -6286,7 +6347,8 @@ static void bfq_completed_request(struct bfq_queue *b= fqq, struct bfq_data *bfqd) =20 bfq_update_hw_tag(bfqd); =20 - bfqd->rq_in_driver--; + bfqd->rq_in_driver[bfqq->actuator_idx]--; + bfqd->tot_rq_in_driver--; bfqq->dispatched--; =20 if (!bfqq->dispatched && !bfq_bfqq_busy(bfqq)) { @@ -6406,7 +6468,7 @@ static void bfq_completed_request(struct bfq_queue *b= fqq, struct bfq_data *bfqd) BFQQE_NO_MORE_REQUESTS); } =20 - if (!bfqd->rq_in_driver) + if (!bfqd->tot_rq_in_driver) bfq_schedule_dispatch(bfqd); } =20 @@ -6537,13 +6599,13 @@ static void bfq_update_inject_limit(struct bfq_data= *bfqd, * conditions to do it, or we can lower the last base value * computed. * - * NOTE: (bfqd->rq_in_driver =3D=3D 1) means that there is no I/O + * NOTE: (bfqd->tot_rq_in_driver =3D=3D 1) means that there is no I/O * request in flight, because this function is in the code * path that handles the completion of a request of bfqq, and, * in particular, this function is executed before - * bfqd->rq_in_driver is decremented in such a code path. + * bfqd->tot_rq_in_driver is decremented in such a code path. */ - if ((bfqq->last_serv_time_ns =3D=3D 0 && bfqd->rq_in_driver =3D=3D 1) || + if ((bfqq->last_serv_time_ns =3D=3D 0 && bfqd->tot_rq_in_driver =3D=3D 1)= || tot_time_ns < bfqq->last_serv_time_ns) { if (bfqq->last_serv_time_ns =3D=3D 0) { /* @@ -6553,7 +6615,7 @@ static void bfq_update_inject_limit(struct bfq_data *= bfqd, bfqq->inject_limit =3D max_t(unsigned int, 1, old_limit); } bfqq->last_serv_time_ns =3D tot_time_ns; - } else if (!bfqd->rqs_injected && bfqd->rq_in_driver =3D=3D 1) + } else if (!bfqd->rqs_injected && bfqd->tot_rq_in_driver =3D=3D 1) /* * No I/O injected and no request still in service in * the drive: these are the exact conditions for @@ -7208,7 +7270,8 @@ static int bfq_init_queue(struct request_queue *q, st= ruct elevator_type *e) bfqd->num_groups_with_pending_reqs =3D 0; #endif =20 - INIT_LIST_HEAD(&bfqd->active_list); + INIT_LIST_HEAD(&bfqd->active_list[0]); + INIT_LIST_HEAD(&bfqd->active_list[1]); INIT_LIST_HEAD(&bfqd->idle_list); INIT_HLIST_HEAD(&bfqd->burst_list); =20 @@ -7253,6 +7316,9 @@ static int bfq_init_queue(struct request_queue *q, st= ruct elevator_type *e) ref_wr_duration[blk_queue_nonrot(bfqd->queue)]; bfqd->peak_rate =3D ref_rate[blk_queue_nonrot(bfqd->queue)] * 2 / 3; =20 + /* see comments on the definition of next field inside bfq_data */ + bfqd->actuator_load_threshold =3D 4; + spin_lock_init(&bfqd->lock); =20 /* diff --git a/block/bfq-iosched.h b/block/bfq-iosched.h index 830dda1f9322..058af701bbbe 100644 --- a/block/bfq-iosched.h +++ b/block/bfq-iosched.h @@ -590,7 +590,12 @@ struct bfq_data { /* number of queued requests */ int queued; /* number of requests dispatched and waiting for completion */ - int rq_in_driver; + int tot_rq_in_driver; + /* + * number of requests dispatched and waiting for completion + * for each actuator + */ + int rq_in_driver[BFQ_MAX_ACTUATORS]; =20 /* true if the device is non rotational and performs queueing */ bool nonrot_with_queueing; @@ -684,8 +689,13 @@ struct bfq_data { /* maximum budget allotted to a bfq_queue before rescheduling */ int bfq_max_budget; =20 - /* list of all the bfq_queues active on the device */ - struct list_head active_list; + /* + * List of all the bfq_queues active for a specific actuator + * on the device. Keeping active queues separate on a + * per-actuator basis helps implementing per-actuator + * injection more efficiently. + */ + struct list_head active_list[BFQ_MAX_ACTUATORS]; /* list of all the bfq_queues idle on the device */ struct list_head idle_list; =20 @@ -821,6 +831,29 @@ struct bfq_data { sector_t sector[BFQ_MAX_ACTUATORS]; sector_t nr_sectors[BFQ_MAX_ACTUATORS]; struct blk_independent_access_range ia_ranges[BFQ_MAX_ACTUATORS]; + + /* + * If the number of I/O requests queued in the device for a + * given actuator is below next threshold, then the actuator + * is deemed as underutilized. If this condition is found to + * hold for some actuator upon a dispatch, but (i) the + * in-service queue does not contain I/O for that actuator, + * while (ii) some other queue does contain I/O for that + * actuator, then the head I/O request of the latter queue is + * returned (injected), instead of the head request of the + * currently in-service queue. + * + * We set the threshold, empirically, to the minimum possible + * value for which an actuator is fully utilized, or close to + * be fully utilized. By doing so, injected I/O 'steals' as + * few drive-queue slots as possibile to the in-service + * queue. This reduces as much as possible the probability + * that the service of I/O from the in-service bfq_queue gets + * delayed because of slot exhaustion, i.e., because all the + * slots of the drive queue are filled with I/O injected from + * other queues (NCQ provides for 32 slots). + */ + unsigned int actuator_load_threshold; }; =20 enum bfqq_state_flags { diff --git a/block/bfq-wf2q.c b/block/bfq-wf2q.c index ea4c3d757fdd..7941b6f07391 100644 --- a/block/bfq-wf2q.c +++ b/block/bfq-wf2q.c @@ -493,7 +493,7 @@ static void bfq_active_insert(struct bfq_service_tree *= st, bfq_update_active_tree(node); =20 if (bfqq) - list_add(&bfqq->bfqq_list, &bfqq->bfqd->active_list); + list_add(&bfqq->bfqq_list, &bfqq->bfqd->active_list[bfqq->actuator_idx]); =20 bfq_inc_active_entities(entity); } --=20 2.20.1 From nobody Tue Sep 16 21:33:18 2025 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 4E75BC53210 for ; Tue, 3 Jan 2023 14:56:09 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237965AbjACO4I (ORCPT ); Tue, 3 Jan 2023 09:56:08 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:60736 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S237976AbjACOzq (ORCPT ); Tue, 3 Jan 2023 09:55:46 -0500 Received: from mail-ej1-x62e.google.com (mail-ej1-x62e.google.com [IPv6:2a00:1450:4864:20::62e]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id E8D531209F for ; Tue, 3 Jan 2023 06:55:34 -0800 (PST) Received: by mail-ej1-x62e.google.com with SMTP id u9so74406297ejo.0 for ; Tue, 03 Jan 2023 06:55:34 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=UgmkxpqJyM32/nscplImRhu54DAKw5l22AaUH8QIVjM=; b=tL2ZtKvOKr8GsSiNzT6HtvXKQzHNy5yXMQchI6xlaOeKOY4LzuvukcFfCFrhSjgsgr qFELHhcsGdq8qqIDyAY48SWFHlpNwvCs0bE/87KGS99zNAUD6yhzuiMcjUaNe35jJnfF 9nYBCkh4GoDV0ZjGmrqoeRbmDCRzH3riaXBQa66FcgeQ1Zp1Cx2wbcnuNZDvs4Z7byyz fcxMlJEfQdgZnLCRK+g3jJy9KO2MwPP2ateNFqqAS7uS76Ol1jxVlNUte34T+07Ox0be I3txtr9ikkxUwgfOHXUnfIXCS4r/gLyZFSCPnbSN5F1YkGOwkERNHFxz5/D19LIIbKvN d0ew== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=UgmkxpqJyM32/nscplImRhu54DAKw5l22AaUH8QIVjM=; b=V/ZN9eW1myN7EqQLhxZIEEVRJETnD60SZUb2VY39LLEbq3fbasGaenOoX/1Ncg/wZp i0SyMpKnQ+fbwZBLH4cEWYGST2Cl7Bm+KQDctP+aSzwB3FZ2lGqLZQmRIyfmExl7ELgF 7ZRRR4UDVkvxtHdTttxY3rD4giUsDUwSnYAy7KhWHypPt0fcmtR44Qeu85bc7N87qq0p VFP8dtEOWQKvy1HDr/tTF0/owQkFsOfnb8zHQAGXEQNOKcn7n/cvzT9xvfLlzc4M/Ee0 5Orr4+xnF/Q38NP8l5iR8O4WddPO+h/zQ/HQiLTpsusePqsc4smarqvYaCD7Y6ZENN+G GFDw== X-Gm-Message-State: AFqh2kqzge5Xyhl6tEVIAeqlP094SRki/JcX2RQtENLztT4FwEK7St8y fuvY/rTE/te8nk1ln+DVimXVCw== X-Google-Smtp-Source: AMrXdXthqpuEe63lXh45UrIXw1Pm+fDkSr40eSGLWJI3kokdasjArLGSfqNYUTQjHK0cWva36m/e7Q== X-Received: by 2002:a17:906:430a:b0:7c1:54b9:c688 with SMTP id j10-20020a170906430a00b007c154b9c688mr37090174ejm.60.1672757734461; Tue, 03 Jan 2023 06:55:34 -0800 (PST) Received: from localhost.localdomain (mob-5-91-46-2.net.vodafone.it. [5.91.46.2]) by smtp.gmail.com with ESMTPSA id g17-20020a17090604d100b007c16fdc93ddsm14122595eja.81.2023.01.03.06.55.33 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Tue, 03 Jan 2023 06:55:34 -0800 (PST) From: Paolo Valente To: Jens Axboe Cc: linux-block@vger.kernel.org, linux-kernel@vger.kernel.org, arie.vanderhoeven@seagate.com, rory.c.chen@seagate.com, glen.valante@linaro.org, damien.lemoal@opensource.wdc.com, Davide Zini , Paolo Valente Subject: [PATCH V13 REBASED 8/8] block, bfq: balance I/O injection among underutilized actuators Date: Tue, 3 Jan 2023 15:55:03 +0100 Message-Id: <20230103145503.71712-9-paolo.valente@linaro.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20230103145503.71712-1-paolo.valente@linaro.org> References: <20230103145503.71712-1-paolo.valente@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" From: Davide Zini Upon the invocation of its dispatch function, BFQ returns the next I/O request of the in-service bfq_queue, unless some exception holds. One such exception is that there is some underutilized actuator, different from the actuator for which the in-service queue contains I/O, and that some other bfq_queue happens to contain I/O for such an actuator. In this case, the next I/O request of the latter bfq_queue, and not of the in-service bfq_queue, is returned (I/O is injected from that bfq_queue). To find such an actuator, a linear scan, in increasing index order, is performed among actuators. Performing a linear scan entails a prioritization among actuators: an underutilized actuator may be considered for injection only if all actuators with a lower index are currently fully utilized, or if there is no pending I/O for any lower-index actuator that happens to be underutilized. This commits breaks this prioritization and tends to distribute injection uniformly across actuators. This is obtained by adding the following condition to the linear scan: even if an actuator A is underutilized, A is however skipped if its load is higher than that of the next actuator. Reviewed-by: Damien Le Moal Signed-off-by: Paolo Valente Signed-off-by: Davide Zini --- block/bfq-iosched.c | 18 +++++++++++++----- 1 file changed, 13 insertions(+), 5 deletions(-) diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c index d42a229b5a86..815b884d6c5a 100644 --- a/block/bfq-iosched.c +++ b/block/bfq-iosched.c @@ -4767,10 +4767,16 @@ bfq_find_active_bfqq_for_actuator(struct bfq_data *= bfqd, int idx) =20 /* * Perform a linear scan of each actuator, until an actuator is found - * for which the following two conditions hold: the load of the - * actuator is below the threshold (see comments on actuator_load_threshold - * for details), and there is a queue that contains I/O for that - * actuator. On success, return that queue. + * for which the following three conditions hold: the load of the + * actuator is below the threshold (see comments on + * actuator_load_threshold for details) and lower than that of the + * next actuator (comments on this extra condition below), and there + * is a queue that contains I/O for that actuator. On success, return + * that queue. + * + * Performing a plain linear scan entails a prioritization among + * actuators. The extra condition above breaks this prioritization and + * tends to distribute injection uniformly across actuators. */ static struct bfq_queue * bfq_find_bfqq_for_underused_actuator(struct bfq_data *bfqd) @@ -4778,7 +4784,9 @@ bfq_find_bfqq_for_underused_actuator(struct bfq_data = *bfqd) int i; =20 for (i =3D 0 ; i < bfqd->num_actuators; i++) { - if (bfqd->rq_in_driver[i] < bfqd->actuator_load_threshold) { + if (bfqd->rq_in_driver[i] < bfqd->actuator_load_threshold && + (i =3D=3D bfqd->num_actuators - 1 || + bfqd->rq_in_driver[i] < bfqd->rq_in_driver[i+1])) { struct bfq_queue *bfqq =3D bfq_find_active_bfqq_for_actuator(bfqd, i); =20 --=20 2.20.1