From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753666Ab3LVDHq (ORCPT ); Sat, 21 Dec 2013 22:07:46 -0500 Received: from moutng.kundenserver.de ([212.227.126.186]:52478 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750898Ab3LVDHp (ORCPT ); Sat, 21 Dec 2013 22:07:45 -0500 Message-ID: <1387681647.5412.25.camel@marge.simpson.net> Subject: Re: [PATCH] rcu: Eliminate softirq processing from rcutree From: Mike Galbraith To: Sebastian Andrzej Siewior Cc: linux-rt-users@vger.kernel.org, Steven Rostedt , linux-kernel@vger.kernel.org, Thomas Gleixner , "Paul E. McKenney" Date: Sun, 22 Dec 2013 04:07:27 +0100 In-Reply-To: <20131221193900.GA8427@linutronix.de> References: <20131221193900.GA8427@linutronix.de> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.2.3 Content-Transfer-Encoding: 7bit Mime-Version: 1.0 X-Provags-ID: V02:K0:6pDV8omx/Mz6yk04mMeUv6Famt9ARw4ckrrtQDPnk5S Vgtdf6JagtRsdAOYlxl/M3WCE6GEOuiYMIYq4QXzAH7734VnrA BYFgjgrJ7OKhNwT0ro7xbzDgKIKv8ZNmVBfmTRrZNBLeCBzRlF 1Il9nvVRTF0WDSrUIdmFSW/v+XL/jKhT1AjpW/LjJKtTl+YvW3 15aWs6e1s26UR/XDCiNXJD+AfCSFoLBIgXsE33XRDYeuX5x0Yd yYw+MdyAB5094+Tjdh9eZWb0NQEqEId27TE/6fL/aVU5MKldoV aDVpwRMYXB5gSKNQqqUFu9hXAfh07YMeujYlGYMy+QmbEZQy8M 8QvV7hAhKL9fAeWT0qGA8vfwgc4EVPZE94W7uOH4FnShvaAcR/ w7KYaW+n9mQ3A== Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sat, 2013-12-21 at 20:39 +0100, Sebastian Andrzej Siewior wrote: > From: "Paul E. McKenney" > > Running RCU out of softirq is a problem for some workloads that would > like to manage RCU core processing independently of other softirq work, > for example, setting kthread priority. This commit therefore moves the > RCU core work from softirq to a per-CPU/per-flavor SCHED_OTHER kthread > named rcuc. The SCHED_OTHER approach avoids the scalability problems > that appeared with the earlier attempt to move RCU core processing to > from softirq to kthreads. That said, kernels built with RCU_BOOST=y > will run the rcuc kthreads at the RCU-boosting priority. I'll take this for a spin on my 64 core test box. I'm pretty sure I'll still end up having to split softirq threads again though, as big box has been unable to meet jitter requirements without, and last upstream rt kernel tested still couldn't. -Mike Hm. Another thing I'll have to check again is btrfs locking fix, and generic IO deadlocks if you don't pull your plug upon first rtmutex block. In 3.0, both were required for box to survive heavy fs pounding. Oh yeah, and the pain of rt tasks playing idle balance for SCHED_OTHER tasks, and nohz balancing crud, and cpupri cost when cores are isolated and and.. sigh, big boxen _suck_ ;-)