From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751321AbWFBIZH (ORCPT ); Fri, 2 Jun 2006 04:25:07 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751324AbWFBIZH (ORCPT ); Fri, 2 Jun 2006 04:25:07 -0400 Received: from mga01.intel.com ([192.55.52.88]:14999 "EHLO fmsmga101-1.fm.intel.com") by vger.kernel.org with ESMTP id S1751321AbWFBIZF (ORCPT ); Fri, 2 Jun 2006 04:25:05 -0400 X-IronPort-AV: i="4.05,202,1146466800"; d="scan'208"; a="45888068:sNHT19185082" From: "Chen, Kenneth W" To: "'Nick Piggin'" , "Con Kolivas" Cc: , "'Chris Mason'" , "Ingo Molnar" Subject: RE: [PATCH RFC] smt nice introduces significant lock contention Date: Fri, 2 Jun 2006 01:24:57 -0700 Message-ID: <000001c6861e$08d426e0$0b4ce984@amr.corp.intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Mailer: Microsoft Office Outlook 11 Thread-Index: AcaF+6loioBfyCuzQlWdmdCTMTmSEAAIiCHg In-Reply-To: <447FBC28.8030401@yahoo.com.au> X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Nick Piggin wrote on Thursday, June 01, 2006 9:19 PM > Con Kolivas wrote: > > On Friday 02 June 2006 12:28, Con Kolivas wrote: > > > >>Actually looking even further, we only introduced the extra lookup of the > >>next task when we started unlocking the runqueue in schedule(). Since we > >>can get by without locking this_rq in schedule with this approach we can > >>simplify dependent_sleeper even further by doing the dependent sleeper > >>check after we have discovered what next is in schedule and avoid looking > >>it up twice. I'll hack something up to do that soon. > > > > > > Something like this (sorry I couldn't help but keep hacking on it). > > Looking pretty good. Nice to acknowledge Chris's idea for > trylocks in your changelog when you submit a final patch. Yes, as far as the lock is concerned in dependent_sleeper(), it looks pretty good. I do have other comments that I will follow up in another thread. - Ken