From mboxrd@z Thu Jan 1 00:00:00 1970 From: Vitaly Kuznetsov Subject: Re: [PATCH 0/2] clocksource/Hyper-V: Add Hyper-V specific sched clock function Date: Mon, 29 Jul 2019 14:13:16 +0200 Message-ID: <87wog1kpib.fsf@vitty.brq.redhat.com> References: <20190729075243.22745-1-Tianyu.Lan@microsoft.com> <87zhkxksxd.fsf@vitty.brq.redhat.com> <20190729110927.GC31398@hirez.programming.kicks-ass.net> Mime-Version: 1.0 Content-Type: text/plain Return-path: In-Reply-To: <20190729110927.GC31398@hirez.programming.kicks-ass.net> Sender: linux-kernel-owner@vger.kernel.org To: Peter Zijlstra Cc: lantianyu1986@gmail.com, Tianyu Lan , linux-arch@vger.kernel.org, linux-hyperv@vger.kernel.org, linux-kernel@vger.kernel.org, luto@kernel.org, tglx@linutronix.de, mingo@redhat.com, bp@alien8.de, hpa@zytor.com, x86@kernel.org, kys@microsoft.com, haiyangz@microsoft.com, sthemmin@microsoft.com, sashal@kernel.org, daniel.lezcano@linaro.org, arnd@arndb.de, michael.h.kelley@microsoft.comashal@kernel.org List-Id: linux-arch.vger.kernel.org Peter Zijlstra writes: > On Mon, Jul 29, 2019 at 12:59:26PM +0200, Vitaly Kuznetsov wrote: >> lantianyu1986@gmail.com writes: >> >> > From: Tianyu Lan >> > >> > Hyper-V guests use the default native_sched_clock() in pv_ops.time.sched_clock >> > on x86. But native_sched_clock() directly uses the raw TSC value, which >> > can be discontinuous in a Hyper-V VM. Add the generic hv_setup_sched_clock() >> > to set the sched clock function appropriately. On x86, this sets >> > pv_ops.time.sched_clock to read the Hyper-V reference TSC value that is >> > scaled and adjusted to be continuous. >> >> Hypervisor can, in theory, disable TSC page and then we're forced to use >> MSR-based clocksource but using it as sched_clock() can be very slow, >> I'm afraid. >> >> On the other hand, what we have now is probably worse: TSC can, >> actually, jump backwards (e.g. on migration) and we're breaking the >> requirements for sched_clock(). > > That (obviously) also breaks the requirements for using TSC as > clocksource. > > IOW, it breaks the entire purpose of having TSC in the first place. Currently, we mark raw TSC as unstable when running on Hyper-V (see 88c9281a9fba6), 'TSC page' (which is TSC * scale + offset) is being used instead. The problem is that 'TSC page' can be disabled by the hypervisor and in that case the only remaining clocksource is MSR-based (slow). -- Vitaly From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-wr1-f65.google.com ([209.85.221.65]:42186 "EHLO mail-wr1-f65.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727003AbfG2MNU (ORCPT ); Mon, 29 Jul 2019 08:13:20 -0400 Received: by mail-wr1-f65.google.com with SMTP id x1so11678641wrr.9 for ; Mon, 29 Jul 2019 05:13:18 -0700 (PDT) From: Vitaly Kuznetsov Subject: Re: [PATCH 0/2] clocksource/Hyper-V: Add Hyper-V specific sched clock function In-Reply-To: <20190729110927.GC31398@hirez.programming.kicks-ass.net> References: <20190729075243.22745-1-Tianyu.Lan@microsoft.com> <87zhkxksxd.fsf@vitty.brq.redhat.com> <20190729110927.GC31398@hirez.programming.kicks-ass.net> Date: Mon, 29 Jul 2019 14:13:16 +0200 Message-ID: <87wog1kpib.fsf@vitty.brq.redhat.com> MIME-Version: 1.0 Content-Type: text/plain Sender: linux-arch-owner@vger.kernel.org List-ID: To: Peter Zijlstra Cc: lantianyu1986@gmail.com, Tianyu Lan , linux-arch@vger.kernel.org, linux-hyperv@vger.kernel.org, linux-kernel@vger.kernel.org, luto@kernel.org, tglx@linutronix.de, mingo@redhat.com, bp@alien8.de, hpa@zytor.com, x86@kernel.org, kys@microsoft.com, haiyangz@microsoft.com, sthemmin@microsoft.com, sashal@kernel.org, daniel.lezcano@linaro.org, arnd@arndb.de, michael.h.kelley@microsoft.comashal@kernel.org Message-ID: <20190729121316.MhQo7-tslXojSIM5B9iOpSiTc_NrAW_4SoimhT_SLcc@z> Peter Zijlstra writes: > On Mon, Jul 29, 2019 at 12:59:26PM +0200, Vitaly Kuznetsov wrote: >> lantianyu1986@gmail.com writes: >> >> > From: Tianyu Lan >> > >> > Hyper-V guests use the default native_sched_clock() in pv_ops.time.sched_clock >> > on x86. But native_sched_clock() directly uses the raw TSC value, which >> > can be discontinuous in a Hyper-V VM. Add the generic hv_setup_sched_clock() >> > to set the sched clock function appropriately. On x86, this sets >> > pv_ops.time.sched_clock to read the Hyper-V reference TSC value that is >> > scaled and adjusted to be continuous. >> >> Hypervisor can, in theory, disable TSC page and then we're forced to use >> MSR-based clocksource but using it as sched_clock() can be very slow, >> I'm afraid. >> >> On the other hand, what we have now is probably worse: TSC can, >> actually, jump backwards (e.g. on migration) and we're breaking the >> requirements for sched_clock(). > > That (obviously) also breaks the requirements for using TSC as > clocksource. > > IOW, it breaks the entire purpose of having TSC in the first place. Currently, we mark raw TSC as unstable when running on Hyper-V (see 88c9281a9fba6), 'TSC page' (which is TSC * scale + offset) is being used instead. The problem is that 'TSC page' can be disabled by the hypervisor and in that case the only remaining clocksource is MSR-based (slow). -- Vitaly