Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752970AbbFXPHM (ORCPT ); Wed, 24 Jun 2015 11:07:12 -0400 Received: from mail-lb0-f172.google.com ([209.85.217.172]:33894 "EHLO mail-lb0-f172.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752266AbbFXPHK (ORCPT ); Wed, 24 Jun 2015 11:07:10 -0400 MIME-Version: 1.0 In-Reply-To: <1435157880-22925-12-git-send-email-riel@redhat.com> References: <1435157880-22925-1-git-send-email-riel@redhat.com> <1435157880-22925-12-git-send-email-riel@redhat.com> From: Andy Lutomirski Date: Wed, 24 Jun 2015 08:06:49 -0700 Message-ID: Subject: Re: [RFC PATCH 11/11] nohz,kvm,time: teach account_process_tick about guest time To: Rik van Riel Cc: "linux-kernel@vger.kernel.org" , fweisbec@redhat.com, Peter Zijlstra , Ingo Molnar , Thomas Gleixner Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 888 Lines: 22 On Wed, Jun 24, 2015 at 7:58 AM, wrote: > From: Rik van Riel > > When tick based accounting is run from a remote CPU, it is actually > possible to encounter a task with PF_VCPU set. Make sure to account > those as guest time. Why do we have PF_VCPU and CONTEXT_GUEST? What's the difference between them (other than the fact that one is per-task and one is per-cpu)? It would be a bit easier to understand if there were fewer of these things. If the issue is that remote sampling would otherwise have a race that could account guest time to the wrong task, then maybe PF_VCPU makes sense. Hmm. --Andy -- 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/