From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id F2A5D10952 for ; Mon, 19 Jun 2023 10:34:39 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 73977C433CA; Mon, 19 Jun 2023 10:34:39 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=linuxfoundation.org; s=korg; t=1687170879; bh=v/K2K951N1AllIDWWWztB0U6RFdbO/hSvfoFqcqQKC0=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=UP6BE1Lj/wV9i8ASl2ZZllcQGmSqjyvORws/Wsn3c8Xzt39POnc383XJGRJlJPkw0 gpnB9wZQseNMo2C5cnVIJrSA6C+KjA80zrKPNcu7pOrjPWNh0/wgAcpGrM9Jgl46hP dzBtKbEJMPWQea7ntLk5gWQoqHO+KUMxU3lLBh8w= From: Greg Kroah-Hartman To: stable@vger.kernel.org Cc: Greg Kroah-Hartman , patches@lists.linux.dev, Kanchan Joshi , Wenwen Chen , Jens Axboe , Sasha Levin Subject: [PATCH 6.3 042/187] io_uring: unlock sqd->lock before sq thread release CPU Date: Mon, 19 Jun 2023 12:27:40 +0200 Message-ID: <20230619102159.648686439@linuxfoundation.org> X-Mailer: git-send-email 2.41.0 In-Reply-To: <20230619102157.579823843@linuxfoundation.org> References: <20230619102157.579823843@linuxfoundation.org> User-Agent: quilt/0.67 Precedence: bulk X-Mailing-List: patches@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: Wenwen Chen [ Upstream commit 533ab73f5b5c95dcb4152b52d5482abcc824c690 ] The sq thread actively releases CPU resources by calling the cond_resched() and schedule() interfaces when it is idle. Therefore, more resources are available for other threads to run. There exists a problem in sq thread: it does not unlock sqd->lock before releasing CPU resources every time. This makes other threads pending on sqd->lock for a long time. For example, the following interfaces all require sqd->lock: io_sq_offload_create(), io_register_iowq_max_workers() and io_ring_exit_work(). Before the sq thread releases CPU resources, unlocking sqd->lock will provide the user a better experience because it can respond quickly to user requests. Signed-off-by: Kanchan Joshi Signed-off-by: Wenwen Chen Link: https://lore.kernel.org/r/20230525082626.577862-1-wenwen.chen@samsung.com Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin --- io_uring/sqpoll.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/io_uring/sqpoll.c b/io_uring/sqpoll.c index 9db4bc1f521a3..5e329e3cd4706 100644 --- a/io_uring/sqpoll.c +++ b/io_uring/sqpoll.c @@ -255,9 +255,13 @@ static int io_sq_thread(void *data) sqt_spin = true; if (sqt_spin || !time_after(jiffies, timeout)) { - cond_resched(); if (sqt_spin) timeout = jiffies + sqd->sq_thread_idle; + if (unlikely(need_resched())) { + mutex_unlock(&sqd->lock); + cond_resched(); + mutex_lock(&sqd->lock); + } continue; } -- 2.39.2