From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-13.5 required=3.0 tests=BAYES_00, DKIM_ADSP_CUSTOM_MED,DKIM_INVALID,DKIM_SIGNED,FREEMAIL_FORGED_FROMDOMAIN, FREEMAIL_FROM,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER, INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 9239AC433E6 for ; Tue, 16 Mar 2021 15:37:33 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id 1DA43650F6 for ; Tue, 16 Mar 2021 15:37:33 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 1DA43650F6 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id AF1938D0015; Tue, 16 Mar 2021 11:37:32 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id ACA158D0014; Tue, 16 Mar 2021 11:37:32 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 8F4BD8D0015; Tue, 16 Mar 2021 11:37:32 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0163.hostedemail.com [216.40.44.163]) by kanga.kvack.org (Postfix) with ESMTP id 70A0D8D0014 for ; Tue, 16 Mar 2021 11:37:32 -0400 (EDT) Received: from smtpin04.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay01.hostedemail.com (Postfix) with ESMTP id 2B1B3180AD837 for ; Tue, 16 Mar 2021 15:37:32 +0000 (UTC) X-FDA: 77926141944.04.E90C4EE Received: from mail-qt1-f178.google.com (mail-qt1-f178.google.com [209.85.160.178]) by imf02.hostedemail.com (Postfix) with ESMTP id 2D5314080F5F for ; Tue, 16 Mar 2021 15:37:30 +0000 (UTC) Received: by mail-qt1-f178.google.com with SMTP id f12so11933510qtq.4 for ; Tue, 16 Mar 2021 08:37:30 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=BJeF+DFkWjXvC2+py6aYx1pM4wZgsYdRz94iG+JwIjQ=; b=GoQJBLW7obUr3v6qda9xFMLeM6y9AGbo4dCEQfj6BGRVxRpRHgk1FfBdowzYTmJ2Xy 2yfzTf8noY/ILScIbrwM2SiENNgJLjQDM1RN6c1Z7F+3KA2If4Twpf2OnqwIEVq5tu0A bZH6CDApzDd1PLtJq/8QqAVkg412e1hs1pK6ugbukBHCzJpAPQ6Z7Uyy+MdOxNJVl8lH In0L+Ion5C8hFvnPRSVJZIOOPdoHM0JsXkLoxaZNjFB2ligzdpragEHG5QS1GfzSsTOo TcHZRWRh/G4mGLlk+C7HUUov7uPPjYj+7UFiEJ/e32N4+T3ngM9aywaLdGe9pOCxhJEt hzzQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=BJeF+DFkWjXvC2+py6aYx1pM4wZgsYdRz94iG+JwIjQ=; b=gxpyGZWYvouyfGC5qJHQcl1ZITsh+3rZPx6cylb2hNypJPk3QI3UmvoeqJRkMRXAVC tUYmq+yGszpKS/1pwPP8eEyf0ao2Zlu1S+pyKWXHp5da4SMrBO6YJQS5DuYbKgkHfW6z xRul+V+ak9n1o6QmVWvjv/nRHhcLygbtJM+SoZG29ZFVdphvzrrlsuMXWX6/BKyCBWDN UInpO/UdQygONhV2OeJiE1tqjDtt507Ixh/YQvH+vfdohJziyVPQLeJv8WPWrM2BtnRM AXk6VWZQJ69jVbD16O12Ek8mjL8Gc78UF3MG/LDEyKx2F1vF+tHBQmGGb1ZdczCUgjd9 LXsQ== X-Gm-Message-State: AOAM533yRNcKzNiXOhtfuwTlxaUocP8aLqs4CUNwatC2gOLFCyV9sdl/ xEc3CFSErzxql7xCdESPV7s= X-Google-Smtp-Source: ABdhPJwuCUat0ipVpAbAZY+Gt6SG/1Qi6JcuoU/MbVmGoUKbejTPsR93MqLGiam1VCeIGV6iNh/e5w== X-Received: by 2002:a05:622a:34f:: with SMTP id r15mr237682qtw.107.1615909040068; Tue, 16 Mar 2021 08:37:20 -0700 (PDT) Received: from localhost ([2620:10d:c091:480::1:1ee]) by smtp.gmail.com with ESMTPSA id k4sm15491144qke.13.2021.03.16.08.37.19 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Tue, 16 Mar 2021 08:37:19 -0700 (PDT) From: Dan Schatzberg To: Cc: Jens Axboe , Tejun Heo , Zefan Li , Johannes Weiner , Andrew Morton , Michal Hocko , Vladimir Davydov , Hugh Dickins , Shakeel Butt , Roman Gushchin , Muchun Song , Alex Shi , Alexander Duyck , Chris Down , Yafang Shao , Wei Yang , linux-block@vger.kernel.org (open list:BLOCK LAYER), linux-kernel@vger.kernel.org (open list), cgroups@vger.kernel.org (open list:CONTROL GROUP (CGROUP)), linux-mm@kvack.org (open list:MEMORY MANAGEMENT) Subject: [PATCH 1/3] loop: Use worker per cgroup instead of kworker Date: Tue, 16 Mar 2021 08:36:50 -0700 Message-Id: <20210316153655.500806-2-schatzberg.dan@gmail.com> X-Mailer: git-send-email 2.30.2 In-Reply-To: <20210316153655.500806-1-schatzberg.dan@gmail.com> References: <20210316153655.500806-1-schatzberg.dan@gmail.com> MIME-Version: 1.0 X-Stat-Signature: kbiy6w3zqm5k3ite4edcsy58gf87z883 X-Rspamd-Server: rspam01 X-Rspamd-Queue-Id: 2D5314080F5F Received-SPF: none (gmail.com>: No applicable sender policy available) receiver=imf02; identity=mailfrom; envelope-from=""; helo=mail-qt1-f178.google.com; client-ip=209.85.160.178 X-HE-DKIM-Result: pass/pass X-HE-Tag: 1615909050-524063 Content-Transfer-Encoding: quoted-printable X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: Existing uses of loop device may have multiple cgroups reading/writing to the same device. Simply charging resources for I/O to the backing file could result in priority inversion where one cgroup gets synchronously blocked, holding up all other I/O to the loop device. In order to avoid this priority inversion, we use a single workqueue where each work item is a "struct loop_worker" which contains a queue of struct loop_cmds to issue. The loop device maintains a tree mapping blk css_id -> loop_worker. This allows each cgroup to independently make forward progress issuing I/O to the backing file. There is also a single queue for I/O associated with the rootcg which can be used in cases of extreme memory shortage where we cannot allocate a loop_worker. The locking for the tree and queues is fairly heavy handed - we acquire a per-loop-device spinlock any time either is accessed. The existing implementation serializes all I/O through a single thread anyways, so I don't believe this is any worse. Signed-off-by: Dan Schatzberg --- drivers/block/loop.c | 207 ++++++++++++++++++++++++++++++++++++------- drivers/block/loop.h | 12 ++- 2 files changed, 182 insertions(+), 37 deletions(-) diff --git a/drivers/block/loop.c b/drivers/block/loop.c index a370cde3ddd4..5c080af73caa 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -70,7 +70,6 @@ #include #include #include -#include #include #include #include @@ -83,6 +82,8 @@ =20 #include =20 +#define LOOP_IDLE_WORKER_TIMEOUT (60 * HZ) + static DEFINE_IDR(loop_index_idr); static DEFINE_MUTEX(loop_ctl_mutex); =20 @@ -920,27 +921,83 @@ static void loop_config_discard(struct loop_device = *lo) q->limits.discard_alignment =3D 0; } =20 -static void loop_unprepare_queue(struct loop_device *lo) -{ - kthread_flush_worker(&lo->worker); - kthread_stop(lo->worker_task); -} +struct loop_worker { + struct rb_node rb_node; + struct work_struct work; + struct list_head cmd_list; + struct list_head idle_list; + struct loop_device *lo; + struct cgroup_subsys_state *css; + unsigned long last_ran_at; +}; =20 -static int loop_kthread_worker_fn(void *worker_ptr) -{ - current->flags |=3D PF_LOCAL_THROTTLE | PF_MEMALLOC_NOIO; - return kthread_worker_fn(worker_ptr); -} +static void loop_workfn(struct work_struct *work); +static void loop_rootcg_workfn(struct work_struct *work); +static void loop_free_idle_workers(struct timer_list *timer); =20 -static int loop_prepare_queue(struct loop_device *lo) +static void loop_queue_work(struct loop_device *lo, struct loop_cmd *cmd= ) { - kthread_init_worker(&lo->worker); - lo->worker_task =3D kthread_run(loop_kthread_worker_fn, - &lo->worker, "loop%d", lo->lo_number); - if (IS_ERR(lo->worker_task)) - return -ENOMEM; - set_user_nice(lo->worker_task, MIN_NICE); - return 0; + struct rb_node **node =3D &(lo->worker_tree.rb_node), *parent =3D NULL; + struct loop_worker *cur_worker, *worker =3D NULL; + struct work_struct *work; + struct list_head *cmd_list; + + spin_lock_irq(&lo->lo_work_lock); + + if (!cmd->css) + goto queue_work; + + node =3D &lo->worker_tree.rb_node; + + while (*node) { + parent =3D *node; + cur_worker =3D container_of(*node, struct loop_worker, rb_node); + if (cur_worker->css =3D=3D cmd->css) { + worker =3D cur_worker; + break; + } else if ((long)cur_worker->css < (long)cmd->css) { + node =3D &(*node)->rb_left; + } else { + node =3D &(*node)->rb_right; + } + } + if (worker) + goto queue_work; + + worker =3D kzalloc(sizeof(struct loop_worker), GFP_NOWAIT | __GFP_NOWAR= N); + /* + * In the event we cannot allocate a worker, just queue on the + * rootcg worker + */ + if (!worker) + goto queue_work; + + worker->css =3D cmd->css; + css_get(worker->css); + INIT_WORK(&worker->work, loop_workfn); + INIT_LIST_HEAD(&worker->cmd_list); + INIT_LIST_HEAD(&worker->idle_list); + worker->lo =3D lo; + rb_link_node(&worker->rb_node, parent, node); + rb_insert_color(&worker->rb_node, &lo->worker_tree); +queue_work: + if (worker) { + /* + * We need to remove from the idle list here while + * holding the lock so that the idle timer doesn't + * free the worker + */ + if (!list_empty(&worker->idle_list)) + list_del_init(&worker->idle_list); + work =3D &worker->work; + cmd_list =3D &worker->cmd_list; + } else { + work =3D &lo->rootcg_work; + cmd_list =3D &lo->rootcg_cmd_list; + } + list_add_tail(&cmd->list_entry, cmd_list); + queue_work(lo->workqueue, work); + spin_unlock_irq(&lo->lo_work_lock); } =20 static void loop_update_rotational(struct loop_device *lo) @@ -1126,12 +1183,27 @@ static int loop_configure(struct loop_device *lo,= fmode_t mode, !file->f_op->write_iter) lo->lo_flags |=3D LO_FLAGS_READ_ONLY; =20 - error =3D loop_prepare_queue(lo); - if (error) + error =3D -EFBIG; + size =3D get_loop_size(lo, file); + if ((loff_t)(sector_t)size !=3D size) goto out_unlock; + lo->workqueue =3D alloc_workqueue("loop%d", + WQ_UNBOUND | WQ_FREEZABLE | + WQ_MEM_RECLAIM, + lo->lo_number); + if (!lo->workqueue) { + error =3D -ENOMEM; + goto out_unlock; + } =20 set_disk_ro(lo->lo_disk, (lo->lo_flags & LO_FLAGS_READ_ONLY) !=3D 0); =20 + INIT_WORK(&lo->rootcg_work, loop_rootcg_workfn); + INIT_LIST_HEAD(&lo->rootcg_cmd_list); + INIT_LIST_HEAD(&lo->idle_worker_list); + lo->worker_tree =3D RB_ROOT; + timer_setup(&lo->timer, loop_free_idle_workers, + TIMER_DEFERRABLE); lo->use_dio =3D lo->lo_flags & LO_FLAGS_DIRECT_IO; lo->lo_device =3D bdev; lo->lo_backing_file =3D file; @@ -1199,6 +1271,7 @@ static int __loop_clr_fd(struct loop_device *lo, bo= ol release) int err =3D 0; bool partscan =3D false; int lo_number; + struct loop_worker *pos, *worker; =20 mutex_lock(&lo->lo_mutex); if (WARN_ON_ONCE(lo->lo_state !=3D Lo_rundown)) { @@ -1218,6 +1291,18 @@ static int __loop_clr_fd(struct loop_device *lo, b= ool release) /* freeze request queue during the transition */ blk_mq_freeze_queue(lo->lo_queue); =20 + destroy_workqueue(lo->workqueue); + spin_lock_irq(&lo->lo_work_lock); + list_for_each_entry_safe(worker, pos, &lo->idle_worker_list, + idle_list) { + list_del(&worker->idle_list); + rb_erase(&worker->rb_node, &lo->worker_tree); + css_put(worker->css); + kfree(worker); + } + spin_unlock_irq(&lo->lo_work_lock); + del_timer_sync(&lo->timer); + spin_lock_irq(&lo->lo_lock); lo->lo_backing_file =3D NULL; spin_unlock_irq(&lo->lo_lock); @@ -1254,7 +1339,6 @@ static int __loop_clr_fd(struct loop_device *lo, bo= ol release) =20 partscan =3D lo->lo_flags & LO_FLAGS_PARTSCAN && bdev; lo_number =3D lo->lo_number; - loop_unprepare_queue(lo); out_unlock: mutex_unlock(&lo->lo_mutex); if (partscan) { @@ -2025,7 +2109,7 @@ static blk_status_t loop_queue_rq(struct blk_mq_hw_= ctx *hctx, } else #endif cmd->css =3D NULL; - kthread_queue_work(&lo->worker, &cmd->work); + loop_queue_work(lo, cmd); =20 return BLK_STS_OK; } @@ -2055,26 +2139,82 @@ static void loop_handle_cmd(struct loop_cmd *cmd) } } =20 -static void loop_queue_work(struct kthread_work *work) +static void loop_set_timer(struct loop_device *lo) +{ + timer_reduce(&lo->timer, jiffies + LOOP_IDLE_WORKER_TIMEOUT); +} + +static void loop_process_work(struct loop_worker *worker, + struct list_head *cmd_list, struct loop_device *lo) { - struct loop_cmd *cmd =3D - container_of(work, struct loop_cmd, work); + int orig_flags =3D current->flags; + struct loop_cmd *cmd; =20 - loop_handle_cmd(cmd); + current->flags |=3D PF_LOCAL_THROTTLE | PF_MEMALLOC_NOIO; + spin_lock_irq(&lo->lo_work_lock); + while (!list_empty(cmd_list)) { + cmd =3D container_of( + cmd_list->next, struct loop_cmd, list_entry); + list_del(cmd_list->next); + spin_unlock_irq(&lo->lo_work_lock); + + loop_handle_cmd(cmd); + cond_resched(); + + spin_lock_irq(&lo->lo_work_lock); + } + + /* + * We only add to the idle list if there are no pending cmds + * *and* the worker will not run again which ensures that it + * is safe to free any worker on the idle list + */ + if (worker && !work_pending(&worker->work)) { + worker->last_ran_at =3D jiffies; + list_add_tail(&worker->idle_list, &lo->idle_worker_list); + loop_set_timer(lo); + } + spin_unlock_irq(&lo->lo_work_lock); + current->flags =3D orig_flags; } =20 -static int loop_init_request(struct blk_mq_tag_set *set, struct request = *rq, - unsigned int hctx_idx, unsigned int numa_node) +static void loop_workfn(struct work_struct *work) { - struct loop_cmd *cmd =3D blk_mq_rq_to_pdu(rq); + struct loop_worker *worker =3D + container_of(work, struct loop_worker, work); + loop_process_work(worker, &worker->cmd_list, worker->lo); +} =20 - kthread_init_work(&cmd->work, loop_queue_work); - return 0; +static void loop_rootcg_workfn(struct work_struct *work) +{ + struct loop_device *lo =3D + container_of(work, struct loop_device, rootcg_work); + loop_process_work(NULL, &lo->rootcg_cmd_list, lo); +} + +static void loop_free_idle_workers(struct timer_list *timer) +{ + struct loop_device *lo =3D container_of(timer, struct loop_device, time= r); + struct loop_worker *pos, *worker; + + spin_lock_irq(&lo->lo_work_lock); + list_for_each_entry_safe(worker, pos, &lo->idle_worker_list, + idle_list) { + if (time_is_after_jiffies(worker->last_ran_at + + LOOP_IDLE_WORKER_TIMEOUT)) + break; + list_del(&worker->idle_list); + rb_erase(&worker->rb_node, &lo->worker_tree); + css_put(worker->css); + kfree(worker); + } + if (!list_empty(&lo->idle_worker_list)) + loop_set_timer(lo); + spin_unlock_irq(&lo->lo_work_lock); } =20 static const struct blk_mq_ops loop_mq_ops =3D { .queue_rq =3D loop_queue_rq, - .init_request =3D loop_init_request, .complete =3D lo_complete_rq, }; =20 @@ -2163,6 +2303,7 @@ static int loop_add(struct loop_device **l, int i) mutex_init(&lo->lo_mutex); lo->lo_number =3D i; spin_lock_init(&lo->lo_lock); + spin_lock_init(&lo->lo_work_lock); disk->major =3D LOOP_MAJOR; disk->first_minor =3D i << part_shift; disk->fops =3D &lo_fops; diff --git a/drivers/block/loop.h b/drivers/block/loop.h index a3c04f310672..9289c1cd6374 100644 --- a/drivers/block/loop.h +++ b/drivers/block/loop.h @@ -14,7 +14,6 @@ #include #include #include -#include #include =20 /* Possible states of device */ @@ -54,8 +53,13 @@ struct loop_device { =20 spinlock_t lo_lock; int lo_state; - struct kthread_worker worker; - struct task_struct *worker_task; + spinlock_t lo_work_lock; + struct workqueue_struct *workqueue; + struct work_struct rootcg_work; + struct list_head rootcg_cmd_list; + struct list_head idle_worker_list; + struct rb_root worker_tree; + struct timer_list timer; bool use_dio; bool sysfs_inited; =20 @@ -66,7 +70,7 @@ struct loop_device { }; =20 struct loop_cmd { - struct kthread_work work; + struct list_head list_entry; bool use_aio; /* use AIO interface to handle I/O */ atomic_t ref; /* only for aio */ long ret; --=20 2.30.2