From patchwork Wed Oct 12 09:40:34 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Yu Kuai X-Patchwork-Id: 13004906 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 2F2C0C43217 for ; Wed, 12 Oct 2022 09:18:35 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229742AbiJLJSb (ORCPT ); Wed, 12 Oct 2022 05:18:31 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:39876 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229677AbiJLJS3 (ORCPT ); Wed, 12 Oct 2022 05:18:29 -0400 Received: from dggsgout11.his.huawei.com (dggsgout11.his.huawei.com [45.249.212.51]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id A8E27BC473; Wed, 12 Oct 2022 02:18:26 -0700 (PDT) Received: from mail02.huawei.com (unknown [172.30.67.153]) by dggsgout11.his.huawei.com (SkyGuard) with ESMTP id 4MnRq05JWpzl911; Wed, 12 Oct 2022 17:16:28 +0800 (CST) Received: from huaweicloud.com (unknown [10.175.127.227]) by APP4 (Coremail) with SMTP id gCh0CgAnKMpehkZjlH+ZAA--.8980S7; Wed, 12 Oct 2022 17:18:25 +0800 (CST) From: Yu Kuai To: tj@kernel.org, axboe@kernel.dk Cc: linux-block@vger.kernel.org, linux-kernel@vger.kernel.org, yukuai3@huawei.com, yukuai1@huaweicloud.com, yi.zhang@huawei.com, linan122@huawei.com Subject: [PATCH v2 3/4] blk-iocost: prevent configuration update concurrent with io throttling Date: Wed, 12 Oct 2022 17:40:34 +0800 Message-Id: <20221012094035.390056-4-yukuai1@huaweicloud.com> X-Mailer: git-send-email 2.31.1 In-Reply-To: <20221012094035.390056-1-yukuai1@huaweicloud.com> References: <20221012094035.390056-1-yukuai1@huaweicloud.com> MIME-Version: 1.0 X-CM-TRANSID: gCh0CgAnKMpehkZjlH+ZAA--.8980S7 X-Coremail-Antispam: 1UD129KBjvJXoWxWF1fXrWUXw4DtFW7Zw15twb_yoW5ArykpF WfGa9xtw40qrs7XFsruF47Xr13W397Kr4xAa97GryfCr17Kr1IqF1vkryj9r48tFZ5Ars8 JFZ5XrWqyry8ArJanT9S1TB71UUUUUUqnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDU0xBIdaVrnRJUUU9m14x267AKxVWrJVCq3wAFc2x0x2IEx4CE42xK8VAvwI8IcIk0 rVWrJVCq3wAFIxvE14AKwVWUJVWUGwA2048vs2IY020E87I2jVAFwI0_JrWl82xGYIkIc2 x26xkF7I0E14v26ryj6s0DM28lY4IEw2IIxxk0rwA2F7IY1VAKz4vEj48ve4kI8wA2z4x0 Y4vE2Ix0cI8IcVAFwI0_tr0E3s1l84ACjcxK6xIIjxv20xvEc7CjxVAFwI0_Gr1j6F4UJw A2z4x0Y4vEx4A2jsIE14v26rxl6s0DM28EF7xvwVC2z280aVCY1x0267AKxVW0oVCq3wAS 0I0E0xvYzxvE52x082IY62kv0487Mc02F40EFcxC0VAKzVAqx4xG6I80ewAv7VC0I7IYx2 IY67AKxVWUJVWUGwAv7VC2z280aVAFwI0_Jr0_Gr1lOx8S6xCaFVCjc4AY6r1j6r4UM4x0 Y48IcxkI7VAKI48JM4x0x7Aq67IIx4CEVc8vx2IErcIFxwCF04k20xvY0x0EwIxGrwCFx2 IqxVCFs4IE7xkEbVWUJVW8JwC20s026c02F40E14v26r1j6r18MI8I3I0E7480Y4vE14v2 6r106r1rMI8E67AF67kF1VAFwI0_Jw0_GFylIxkGc2Ij64vIr41lIxAIcVC0I7IYx2IY67 AKxVWUJVWUCwCI42IY6xIIjxv20xvEc7CjxVAFwI0_Gr0_Cr1lIxAIcVCF04k26cxKx2IY s7xG6r1j6r1xMIIF0xvEx4A2jsIE14v26r1j6r4UMIIF0xvEx4A2jsIEc7CjxVAFwI0_Gr 0_Gr1UYxBIdaVFxhVjvjDU0xZFpf9x0JUd8n5UUUUU= X-CM-SenderInfo: 51xn3trlr6x35dzhxuhorxvhhfrp/ X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-block@vger.kernel.org From: Yu Kuai This won't cause any severe problem currently, however, this doesn't seems appropriate: 1) 'ioc->params' is read from multiple places without holding 'ioc->lock', unexpected value might be read if writing it concurrently. 2) If configuration is changed while io is throttling, the functionality might be affected. For example, if module params is updated and cost becomes smaller, waiting for timer that is caculated under old configuration is not appropriate. Signed-off-by: Yu Kuai Acked-by: Tejun Heo --- block/blk-iocost.c | 26 ++++++++++++++++++++++++-- 1 file changed, 24 insertions(+), 2 deletions(-) diff --git a/block/blk-iocost.c b/block/blk-iocost.c index 6d36a4bd4382..5acc5f13bbd6 100644 --- a/block/blk-iocost.c +++ b/block/blk-iocost.c @@ -3187,6 +3187,9 @@ static ssize_t ioc_qos_write(struct kernfs_open_file *of, char *input, ioc = q_to_ioc(disk->queue); } + blk_mq_freeze_queue(disk->queue); + blk_mq_quiesce_queue(disk->queue); + spin_lock_irq(&ioc->lock); memcpy(qos, ioc->params.qos, sizeof(qos)); enable = ioc->enabled; @@ -3278,10 +3281,17 @@ static ssize_t ioc_qos_write(struct kernfs_open_file *of, char *input, ioc_refresh_params(ioc, true); spin_unlock_irq(&ioc->lock); + blk_mq_unquiesce_queue(disk->queue); + blk_mq_unfreeze_queue(disk->queue); + blkdev_put_no_open(bdev); return nbytes; einval: spin_unlock_irq(&ioc->lock); + + blk_mq_unquiesce_queue(disk->queue); + blk_mq_unfreeze_queue(disk->queue); + ret = -EINVAL; err: blkdev_put_no_open(bdev); @@ -3336,6 +3346,7 @@ static ssize_t ioc_cost_model_write(struct kernfs_open_file *of, char *input, size_t nbytes, loff_t off) { struct block_device *bdev; + struct request_queue *q; struct ioc *ioc; u64 u[NR_I_LCOEFS]; bool user; @@ -3346,14 +3357,18 @@ static ssize_t ioc_cost_model_write(struct kernfs_open_file *of, char *input, if (IS_ERR(bdev)) return PTR_ERR(bdev); - ioc = q_to_ioc(bdev_get_queue(bdev)); + q = bdev_get_queue(bdev); + ioc = q_to_ioc(q); if (!ioc) { ret = blk_iocost_init(bdev->bd_disk); if (ret) goto err; - ioc = q_to_ioc(bdev_get_queue(bdev)); + ioc = q_to_ioc(q); } + blk_mq_freeze_queue(q); + blk_mq_quiesce_queue(q); + spin_lock_irq(&ioc->lock); memcpy(u, ioc->params.i_lcoefs, sizeof(u)); user = ioc->user_cost_model; @@ -3402,11 +3417,18 @@ static ssize_t ioc_cost_model_write(struct kernfs_open_file *of, char *input, ioc_refresh_params(ioc, true); spin_unlock_irq(&ioc->lock); + blk_mq_unquiesce_queue(q); + blk_mq_unfreeze_queue(q); + blkdev_put_no_open(bdev); return nbytes; einval: spin_unlock_irq(&ioc->lock); + + blk_mq_unquiesce_queue(q); + blk_mq_unfreeze_queue(q); + ret = -EINVAL; err: blkdev_put_no_open(bdev);