Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754338Ab3IXJcr (ORCPT ); Tue, 24 Sep 2013 05:32:47 -0400 Received: from gate.crashing.org ([63.228.1.57]:55048 "EHLO gate.crashing.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753868Ab3IXJcq (ORCPT ); Tue, 24 Sep 2013 05:32:46 -0400 Message-ID: <1380015099.5443.43.camel@pasglop> Subject: Re: [RFC GIT PULL] softirq: Consolidation and stack overrun fix From: Benjamin Herrenschmidt To: Peter Zijlstra Cc: Linus Torvalds , "H. Peter Anvin" , Frederic Weisbecker , Thomas Gleixner , LKML , Paul Mackerras , Ingo Molnar , James Hogan , "James E.J. Bottomley" , Helge Deller , Martin Schwidefsky , Heiko Carstens , "David S. Miller" , Andrew Morton , Anton Blanchard Date: Tue, 24 Sep 2013 19:31:39 +1000 In-Reply-To: <20130924082128.GJ9326@twins.programming.kicks-ass.net> References: <1379824754.24090.11.camel@pasglop> <1379824861.24090.12.camel@pasglop> <20130922162410.GA10649@laptop.programming.kicks-ass.net> <1379887000.24090.19.camel@pasglop> <1379981427.5443.8.camel@pasglop> <1379987527.5443.20.camel@pasglop> <20130924080436.GH9326@twins.programming.kicks-ass.net> <1380010613.5443.40.camel@pasglop> <20130924082128.GJ9326@twins.programming.kicks-ass.net> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.6.4-0ubuntu1 Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1078 Lines: 26 On Tue, 2013-09-24 at 10:21 +0200, Peter Zijlstra wrote: > > I don't see how that relates to the above though... > > It was a comment on the increase of per-cpu usage in generic code. Oh I see. Ok, but some things are intrinsically per-task so if you use per-cpu in that case you end up context switching, at least that's what Linus alluded to. It would be nice if such construct we design so that we still access them via the task struct on arch where such accesses are cheap. I think if/when we get rid of using r13 for our arch specific "PACA", it becomes easy for me to experiment with using it for either per-cpu or current and do some measurements. It also means I can easily switch "the other way" if I need to. I'll dig into that, it will probably take some time before I sanitize our code enough anyway. Cheers, Ben. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/