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 mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 3DAA1C433FE for ; Mon, 15 Nov 2021 18:44:57 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 20D54635DA for ; Mon, 15 Nov 2021 18:44:57 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S242448AbhKOSrq (ORCPT ); Mon, 15 Nov 2021 13:47:46 -0500 Received: from mail.kernel.org ([198.145.29.99]:52664 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S238113AbhKORgG (ORCPT ); Mon, 15 Nov 2021 12:36:06 -0500 Received: by mail.kernel.org (Postfix) with ESMTPSA id 581B460EE9; Mon, 15 Nov 2021 17:24:06 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=linuxfoundation.org; s=korg; t=1636997046; bh=uD1fB78HhVPIW6Ft/n8J06SJfAHrcJhtXKN5Z4rvx9E=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=2Ib8MKmFItBSpwSgJgxcbezkZgjYs5ajurvqu9agc36RYbgUfvjvDbmNJDq7VK7SQ nX88eKz4mDwOiYmXZN2dm/x1JoQdBaOsryZLUOwFPGFARyeiGlaDlC6478aCGw2VSH fJEiCuxfSnU2DBSD31v/DanE0OzEnGKJ8p+xXfRc= From: Greg Kroah-Hartman To: linux-kernel@vger.kernel.org Cc: Greg Kroah-Hartman , stable@vger.kernel.org, Trond Myklebust Subject: [PATCH 5.4 355/355] SUNRPC: Partial revert of commit 6f9f17287e78 Date: Mon, 15 Nov 2021 18:04:39 +0100 Message-Id: <20211115165325.232105972@linuxfoundation.org> X-Mailer: git-send-email 2.33.1 In-Reply-To: <20211115165313.549179499@linuxfoundation.org> References: <20211115165313.549179499@linuxfoundation.org> User-Agent: quilt/0.66 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org From: Trond Myklebust commit ea7a1019d8baf8503ecd6e3ec8436dec283569e6 upstream. The premise of commit 6f9f17287e78 ("SUNRPC: Mitigate cond_resched() in xprt_transmit()") was that cond_resched() is expensive and unnecessary when there has been just a single send. The point of cond_resched() is to ensure that tasks that should pre-empt this one get a chance to do so when it is safe to do so. The code prior to commit 6f9f17287e78 failed to take into account that it was keeping a rpc_task pinned for longer than it needed to, and so rather than doing a full revert, let's just move the cond_resched. Signed-off-by: Trond Myklebust Signed-off-by: Greg Kroah-Hartman --- net/sunrpc/xprt.c | 28 +++++++++++++++------------- 1 file changed, 15 insertions(+), 13 deletions(-) --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -1538,15 +1538,14 @@ xprt_transmit(struct rpc_task *task) { struct rpc_rqst *next, *req = task->tk_rqstp; struct rpc_xprt *xprt = req->rq_xprt; - int counter, status; + int status; spin_lock(&xprt->queue_lock); - counter = 0; - while (!list_empty(&xprt->xmit_queue)) { - if (++counter == 20) + for (;;) { + next = list_first_entry_or_null(&xprt->xmit_queue, + struct rpc_rqst, rq_xmit); + if (!next) break; - next = list_first_entry(&xprt->xmit_queue, - struct rpc_rqst, rq_xmit); xprt_pin_rqst(next); spin_unlock(&xprt->queue_lock); status = xprt_request_transmit(next, task); @@ -1554,13 +1553,16 @@ xprt_transmit(struct rpc_task *task) status = 0; spin_lock(&xprt->queue_lock); xprt_unpin_rqst(next); - if (status == 0) { - if (!xprt_request_data_received(task) || - test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) - continue; - } else if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) - task->tk_status = status; - break; + if (status < 0) { + if (test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) + task->tk_status = status; + break; + } + /* Was @task transmitted, and has it received a reply? */ + if (xprt_request_data_received(task) && + !test_bit(RPC_TASK_NEED_XMIT, &task->tk_runstate)) + break; + cond_resched_lock(&xprt->queue_lock); } spin_unlock(&xprt->queue_lock); }