Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
        id S1752111AbcLLLl0 (ORCPT <rfc822;w@1wt.eu>);
        Mon, 12 Dec 2016 06:41:26 -0500
Received: from dcvr.yhbt.net ([64.71.152.64]:56368 "EHLO dcvr.yhbt.net"
        rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
        id S1750933AbcLLLlY (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
        Mon, 12 Dec 2016 06:41:24 -0500
X-Greylist: delayed 441 seconds by postgrey-1.27 at vger.kernel.org; Mon, 12 Dec 2016 06:41:24 EST
Date: Mon, 12 Dec 2016 11:34:01 +0000
From: Eric Wong <e@80x24.org>
To: Dmitry Banschikov <me@ubique.spb.ru>
Cc: linux-kernel@vger.kernel.org, John Stultz <john.stultz@linaro.org>,
        Deepa Dinamani <deepa.kernel@gmail.com>
Subject: Re: epoll_wait inaccurate timeout
Message-ID: <20161212113401.GA25817@starla>
References: <CABuNDZTctuL_m31bYTK5XjrEO-CtYM07xOBoZWBcgcTgM5qE_Q@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <CABuNDZTctuL_m31bYTK5XjrEO-CtYM07xOBoZWBcgcTgM5qE_Q@mail.gmail.com>
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org
Content-Length: 952
Lines: 30

+Cc folks who may know about timer stuff on epoll.

Dmitry Banschikov <me@ubique.spb.ru> wrote:
> Hi!
> 
> I have a problem caused by inaccurate timeouts in epoll_wait(2).
> Here are some parts of strace -tt output:

Which kernel version are you using?

> 22578 09:33:46.959791 epoll_wait(5,  <unfinished ...>
> 22578 09:33:50.010794 <... epoll_wait resumed> [], 128, 1498) = 0
> ...
> 22034 09:35:07.686896 epoll_wait(5,  <unfinished ...>
> 22034 09:35:09.482526 <... epoll_wait resumed> [{EPOLLIN,
> {u32=151458248, u64=151458248}}], 128, 362) = 1
> ...
> 22036 09:35:41.433241 epoll_wait(5,  <unfinished ...>
> 22036 09:35:43.176881 <... epoll_wait resumed> [], 128, 97) = 0
> 
> In each example epoll_wait is blocked for too longer then asked in timeout.
> 
> Is it normal?

I don't think so, unless you have a huge /proc/<pid>/timerslack_ns
set.  But I mainly use -1 or 0 as the timeout value.

> Please CC me.

That's standard procedure, here :)