From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mathieu Desnoyers Subject: Re: [RFC PATCH 4/7] x86: use exit_lazy_tlb rather than membarrier_mm_sync_core_before_usermode Date: Fri, 17 Jul 2020 12:22:49 -0400 (EDT) Message-ID: <12700909.18968.1595002969773.JavaMail.zimbra@efficios.com> References: <20200710015646.2020871-1-npiggin@gmail.com> <1370747990.15974.1594915396143.JavaMail.zimbra@efficios.com> <595582123.17106.1594925921537.JavaMail.zimbra@efficios.com> <20200716212416.GA1126458@rowland.harvard.edu> <1770378591.18523.1594993165391.JavaMail.zimbra@efficios.com> <20200717145102.GC1147780@rowland.harvard.edu> <1697220787.18880.1595000348405.JavaMail.zimbra@efficios.com> <20200717161145.GA1150454@rowland.harvard.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20200717161145.GA1150454@rowland.harvard.edu> Sender: linux-kernel-owner@vger.kernel.org To: Alan Stern Cc: Nicholas Piggin , paulmck , Anton Blanchard , Arnd Bergmann , linux-arch , linux-kernel , linux-mm , linuxppc-dev , Andy Lutomirski , Peter Zijlstra , x86 List-Id: linux-arch.vger.kernel.org ----- On Jul 17, 2020, at 12:11 PM, Alan Stern stern@rowland.harvard.edu wrote: >> > I agree with Nick: A memory barrier is needed somewhere between the >> > assignment at 6 and the return to user mode at 8. Otherwise you end up >> > with the Store Buffer pattern having a memory barrier on only one side, >> > and it is well known that this arrangement does not guarantee any >> > ordering. >> >> Yes, I see this now. I'm still trying to wrap my head around why the memory >> barrier at the end of membarrier() needs to be paired with a scheduler >> barrier though. > > The memory barrier at the end of membarrier() on CPU0 is necessary in > order to enforce the guarantee that any writes occurring on CPU1 before > the membarrier() is executed will be visible to any code executing on > CPU0 after the membarrier(). Ignoring the kthread issue, we can have: > > CPU0 CPU1 > x = 1 > barrier() > y = 1 > r2 = y > membarrier(): > a: smp_mb() > b: send IPI IPI-induced mb > c: smp_mb() > r1 = x > > The writes to x and y are unordered by the hardware, so it's possible to > have r2 = 1 even though the write to x doesn't execute until b. If the > memory barrier at c is omitted then "r1 = x" can be reordered before b > (although not before a), so we get r1 = 0. This violates the guarantee > that membarrier() is supposed to provide. > > The timing of the memory barrier at c has to ensure that it executes > after the IPI-induced memory barrier on CPU1. If it happened before > then we could still end up with r1 = 0. That's why the pairing matters. > > I hope this helps your head get properly wrapped. :-) It does help a bit! ;-) This explains this part of the comment near the smp_mb at the end of membarrier: * Memory barrier on the caller thread _after_ we finished * waiting for the last IPI. [...] However, it does not explain why it needs to be paired with a barrier in the scheduler, clearly for the case where the IPI is skipped. I wonder whether this part of the comment is factually correct: * [...] Matches memory barriers around rq->curr modification in scheduler. Thanks, Mathieu -- Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Date: Fri, 17 Jul 2020 12:22:49 -0400 (EDT) From: Mathieu Desnoyers Message-ID: <12700909.18968.1595002969773.JavaMail.zimbra@efficios.com> In-Reply-To: <20200717161145.GA1150454@rowland.harvard.edu> References: <20200710015646.2020871-1-npiggin@gmail.com> <1370747990.15974.1594915396143.JavaMail.zimbra@efficios.com> <595582123.17106.1594925921537.JavaMail.zimbra@efficios.com> <20200716212416.GA1126458@rowland.harvard.edu> <1770378591.18523.1594993165391.JavaMail.zimbra@efficios.com> <20200717145102.GC1147780@rowland.harvard.edu> <1697220787.18880.1595000348405.JavaMail.zimbra@efficios.com> <20200717161145.GA1150454@rowland.harvard.edu> Subject: Re: [RFC PATCH 4/7] x86: use exit_lazy_tlb rather than membarrier_mm_sync_core_before_usermode MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org To: Alan Stern Cc: Nicholas Piggin , paulmck , Anton Blanchard , Arnd Bergmann , linux-arch , linux-kernel , linux-mm , linuxppc-dev , Andy Lutomirski , Peter Zijlstra , x86 List-ID: Message-ID: <20200717162249.FcMlAfDN83uYHTBEW_sYi1L8LSzl_wcAEmbQfey7Cyk@z> ----- On Jul 17, 2020, at 12:11 PM, Alan Stern stern@rowland.harvard.edu wrote: >> > I agree with Nick: A memory barrier is needed somewhere between the >> > assignment at 6 and the return to user mode at 8. Otherwise you end up >> > with the Store Buffer pattern having a memory barrier on only one side, >> > and it is well known that this arrangement does not guarantee any >> > ordering. >> >> Yes, I see this now. I'm still trying to wrap my head around why the memory >> barrier at the end of membarrier() needs to be paired with a scheduler >> barrier though. > > The memory barrier at the end of membarrier() on CPU0 is necessary in > order to enforce the guarantee that any writes occurring on CPU1 before > the membarrier() is executed will be visible to any code executing on > CPU0 after the membarrier(). Ignoring the kthread issue, we can have: > > CPU0 CPU1 > x = 1 > barrier() > y = 1 > r2 = y > membarrier(): > a: smp_mb() > b: send IPI IPI-induced mb > c: smp_mb() > r1 = x > > The writes to x and y are unordered by the hardware, so it's possible to > have r2 = 1 even though the write to x doesn't execute until b. If the > memory barrier at c is omitted then "r1 = x" can be reordered before b > (although not before a), so we get r1 = 0. This violates the guarantee > that membarrier() is supposed to provide. > > The timing of the memory barrier at c has to ensure that it executes > after the IPI-induced memory barrier on CPU1. If it happened before > then we could still end up with r1 = 0. That's why the pairing matters. > > I hope this helps your head get properly wrapped. :-) It does help a bit! ;-) This explains this part of the comment near the smp_mb at the end of membarrier: * Memory barrier on the caller thread _after_ we finished * waiting for the last IPI. [...] However, it does not explain why it needs to be paired with a barrier in the scheduler, clearly for the case where the IPI is skipped. I wonder whether this part of the comment is factually correct: * [...] Matches memory barriers around rq->curr modification in scheduler. Thanks, Mathieu -- Mathieu Desnoyers EfficiOS Inc. http://www.efficios.com