Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Wed, 28 Aug 2002 13:08:57 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Wed, 28 Aug 2002 13:08:57 -0400 Received: from mx1.elte.hu ([157.181.1.137]:42467 "HELO mx1.elte.hu") by vger.kernel.org with SMTP id ; Wed, 28 Aug 2002 13:08:56 -0400 Date: Wed, 28 Aug 2002 19:16:31 +0200 (CEST) From: Ingo Molnar Reply-To: Ingo Molnar To: Rusty Russell Cc: linux-kernel@vger.kernel.org, Linus Torvalds Subject: Re: [patch] "fully HT-aware scheduler" support, 2.5.31-BK-curr In-Reply-To: <20020828163242.2c84747f.rusty@rustcorp.com.au> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1243 Lines: 32 On Wed, 28 Aug 2002, Rusty Russell wrote: > > - HT-aware affinity. > > > > Tasks should attempt to 'stick' to physical CPUs, not logical CPUs. > > Linus disagreed with this before when I discussed it with him, and with > the current (stupid, non-portable, broken) set_affinity syscall he's > right. actually, affinity still works just fine, users can bind tasks to logical CPUs as well. What i meant was the affinity logic of the scheduler (ie. affinity decisions done by the scheduler), not the externally visible affinity API. > You don't know if someone said "schedule me on cpu 0" because they > really want to be scheduled on CPU 0, or because they really *don't* > want to be scheduled on CPU 1 (where something else is running). You > can't just assume they are equivalent if they are the same physical CPU. i dont assume that. There's also a fair amount of code in the kernel that relies on binding threads to particular CPUs, the patch does not break that in any way. Ingo - 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/