Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262515AbVDLRWN (ORCPT ); Tue, 12 Apr 2005 13:22:13 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262471AbVDLRRJ (ORCPT ); Tue, 12 Apr 2005 13:17:09 -0400 Received: from palrel12.hp.com ([156.153.255.237]:30390 "EHLO palrel12.hp.com") by vger.kernel.org with ESMTP id S262464AbVDLRP7 (ORCPT ); Tue, 12 Apr 2005 13:15:59 -0400 From: David Mosberger MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <16988.586.986381.393504@napali.hpl.hp.com> Date: Tue, 12 Apr 2005 10:15:54 -0700 To: Ingo Molnar Cc: davidm@hpl.hp.com, "David S. Miller" , tony.luck@intel.com, linux-kernel@vger.kernel.org Subject: Re: [patch] sched: unlocked context-switches In-Reply-To: <20050412064253.GA1289@elte.hu> References: <3R6Ir-89Y-23@gated-at.bofh.it> <20050409070738.GA5444@elte.hu> <16983.33049.962002.335198@napali.hpl.hp.com> <20050409155810.593d8f7b.davem@davemloft.net> <20050410064324.GA24596@elte.hu> <16987.7956.806699.617633@napali.hpl.hp.com> <20050412064253.GA1289@elte.hu> X-Mailer: VM 7.19 under Emacs 21.4.1 Reply-To: davidm@hpl.hp.com X-URL: http://www.hpl.hp.com/personal/David_Mosberger/ Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1293 Lines: 32 >>>>> On Tue, 12 Apr 2005 08:42:53 +0200, Ingo Molnar said: Ingo> * David Mosberger wrote: >> Now, Ingo says that the order is reversed with his patch, i.e., >> switch_mm() happens after switch_to(). That means flush_tlb_mm() >> may now see a current->active_mm which hasn't really been >> activated yet. That should be OK since it would just mean that >> we'd do an early (and duplicate) activate_context(). While it >> does not give me a warm and fuzzy feeling to have this >> inconsistent state be observable by interrupt-handlers (and, in >> particular, IPI-handlers), I don't see any problem with it off >> hand. Ingo> thanks for the analysis. I fundamentally dont have any fuzzy Ingo> feeling from having _any_ portion of the context-switch path Ingo> nonatomic, but with more complex hardware it's just not Ingo> possible it seems. No kidding! ;-) I _think_ the change is OK. I'll need testing, of course. Sure would be nice to have 2.7.xx... Thanks, --david - 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/