Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751610AbaL0JDg (ORCPT ); Sat, 27 Dec 2014 04:03:36 -0500 Received: from szxga01-in.huawei.com ([119.145.14.64]:20435 "EHLO szxga01-in.huawei.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751079AbaL0JD1 (ORCPT ); Sat, 27 Dec 2014 04:03:27 -0500 Message-ID: <549E75C3.80808@huawei.com> Date: Sat, 27 Dec 2014 17:02:59 +0800 From: Li Bin User-Agent: Mozilla/5.0 (Windows NT 6.1; rv:24.0) Gecko/20100101 Thunderbird/24.0.1 MIME-Version: 1.0 To: Sasha Levin , Peter Zijlstra CC: Ingo Molnar , LKML , "Dave Jones" , , Subject: Re: sched: spinlock recursion in sched_rr_get_interval References: <53B98709.3090603@oracle.com> <20140707083016.GA19379@twins.programming.kicks-ass.net> <53BAA6DF.5060409@oracle.com> <20140707200550.GA6758@twins.programming.kicks-ass.net> <549D03F6.9090607@huawei.com> <549D07CD.8000204@oracle.com> In-Reply-To: <549D07CD.8000204@oracle.com> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 8bit X-Originating-IP: [10.177.25.181] X-CFilter-Loop: Reflected Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 2014/12/26 15:01, Sasha Levin wrote: > On 12/26/2014 01:45 AM, Li Bin wrote: >> On 2014/7/8 4:05, Peter Zijlstra wrote: >>>> On Mon, Jul 07, 2014 at 09:55:43AM -0400, Sasha Levin wrote: >>>>>> I've also had this one, which looks similar: >>>>>> >>>>>> [10375.005884] BUG: spinlock recursion on CPU#0, modprobe/10965 >>>>>> [10375.006573] lock: 0xffff8803a0fd7740, .magic: dead4ead, .owner: modprobe/10965, .owner_cpu: 15 >>>>>> [10375.007412] CPU: 0 PID: 10965 Comm: modprobe Tainted: G W 3.16.0-rc3-next-20140704-sasha-00023-g26c0906-dirty #765 >>>> >>>> Something's fucked; so we have: >>>> >>>> debug_spin_lock_before() >>>> SPIN_BUG_ON(lock->owner == current, "recursion"); >>>> >> Hello, >> Does ACCESS_ONCE() can help this issue? I have no evidence that its lack is >> responsible for the issue, but I think here need it indeed. Is that right? >> >> SPIN_BUG_ON(ACCESS_ONCE(lock->owner) == current, "recursion"); > > Could you explain a bit more why do you think it's needed? > Oh, just adding ACCESS_ONCE may be not enough, and i think lacking lock protection for reading lock->owner is a risk. In short, the reason of the issue is more like the spinlock debug mechanism, rather than a real spinlock recursion. ... //under no lock protection if (lock->owner == current) //access lock->owner |-spin_dump(lock, "recursion"); |-if (lock->owner && lock->owner != SPINLOCK_OWNER_INIT) //access lock->owner again owner = lock->owner; ... Right, or am I missing something? Thanks, Li Bin > > Thanks, > Sasha > > -- 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/