Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752208AbdF1Qqv (ORCPT ); Wed, 28 Jun 2017 12:46:51 -0400 Received: from mail-yw0-f194.google.com ([209.85.161.194]:34066 "EHLO mail-yw0-f194.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752085AbdF1Qqp (ORCPT ); Wed, 28 Jun 2017 12:46:45 -0400 MIME-Version: 1.0 In-Reply-To: <20170628101248.GB5981@leverpostej> References: <2256f9b5-1277-c4b1-1472-61a10cd1db9a@linux.intel.com> <20170628101248.GB5981@leverpostej> From: Kyle Huey Date: Wed, 28 Jun 2017 09:46:43 -0700 Message-ID: Subject: Re: [REGRESSION] perf/core: PMU interrupts dropped if we entered the kernel in the "skid" region To: Mark Rutland Cc: "Jin, Yao" , Ingo Molnar , "Peter Zijlstra (Intel)" , stable@vger.kernel.org, Alexander Shishkin , Arnaldo Carvalho de Melo , Jiri Olsa , Linus Torvalds , Namhyung Kim , Stephane Eranian , Thomas Gleixner , Vince Weaver , acme@kernel.org, jolsa@kernel.org, kan.liang@intel.com, Will Deacon , gregkh@linuxfoundation.org, "Robert O'Callahan" , open list 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: 2727 Lines: 65 On Wed, Jun 28, 2017 at 3:12 AM, Mark Rutland wrote: > On Tue, Jun 27, 2017 at 09:51:00PM -0700, Kyle Huey wrote: >> On Tue, Jun 27, 2017 at 7:09 PM, Jin, Yao wrote: >> > Hi, >> > >> > In theory, the PMI interrupts in skid region should be dropped, right? >> >> No, why would they be dropped? >> >> My understanding of the situation is as follows: >> >> There is some time, call it t_0, where the hardware counter overflows. >> The PMU triggers an interrupt, but this is not instantaneous. Call >> the time when the interrupt is actually delivered t_1. Then t_1 - t_0 >> is the "skid". >> >> Note that if the counter is `exclude_kernel`, then at t_0 the CPU >> *must* be running a userspace program. But by t_1, the CPU may be >> doing something else. Your patch changed things so that if at t_1 the >> CPU is in the kernel, then the interrupt is discarded. But rr has >> programmed the counter to deliver a signal on overflow (via F_SETSIG >> on the fd returned by perf_event_open). This change results in the >> signal never being delivered, because the interrupt was ignored. >> (More accurately, the signal is delivered the *next* time the counter >> overflows, which is far past where we wanted to inject our >> asynchronous event into our tracee. > > Yes, this is a bug. > > As we're trying to avoid smapling state, I think we can move the check > into perf_prepare_sample() or __perf_event_output(), where that state is > actually sampled. I'll take a look at that momentarily. > > Just to clarify, you don't care about the sample state at all? i.e. you > don't need the user program counter? Right. `sample_regs_user`, `sample_star_user`, `branch_sample_type`, etc are all 0. https://github.com/mozilla/rr/blob/cf594dd01f07d96a61409e9f41a29f78c8c51693/src/PerfCounters.cc#L194 is what we do use. > Is that signal delivered to the tracee, or to a different process that > traces it? If the latter, what ensures that the task is stopped > sufficiently quickly? It's delivered to the tracee (via an F_SETOWN_EX with the tracee tid). In practice we've found that on modern Intel hardware that the interrupt and resulting signal delivery delay is bounded by a relatively small number of counter events. >> It seems to me that it might be reasonable to ignore the interrupt if >> the purpose of the interrupt is to trigger sampling of the CPUs >> register state. But if the interrupt will trigger some other >> operation, such as a signal on an fd, then there's no reason to drop >> it. > > Agreed. I'll try to have a patch for this soon. > > I just need to figure out exactly where that overflow signal is > generated by the perf core. > > Thanks, > Mark. - Kyle