From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jens Axboe Subject: Re: [PATCH 1/11] Add generic helpers for arch IPI function calls Date: Wed, 23 Apr 2008 09:49:33 +0200 Message-ID: <20080423074933.GB12774@kernel.dk> References: <1208890227-24808-1-git-send-email-jens.axboe@oracle.com> <1208890227-24808-2-git-send-email-jens.axboe@oracle.com> <1208895423.7115.290.camel@twins> <20080423060725.GT12774@kernel.dk> <1208932320.7115.319.camel@twins> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <1208932320.7115.319.camel@twins> Sender: linux-arch-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org List-ID: To: Peter Zijlstra Cc: linux-arch-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, npiggin-l3A5Bk7waGM@public.gmane.org, torvalds-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org, sam-uyr5N9Q2VtJg9hUCZPvPmw@public.gmane.org On Wed, Apr 23 2008, Peter Zijlstra wrote: > On Wed, 2008-04-23 at 08:07 +0200, Jens Axboe wrote: > > On Tue, Apr 22 2008, Peter Zijlstra wrote: > > > On Tue, 2008-04-22 at 20:50 +0200, Jens Axboe wrote: > > > > > > > +int smp_call_function_single(int cpu, void (*func) (void *info), void *info, > > > > + int retry, int wait) > > > > +{ > > > > + unsigned long flags; > > > > + /* prevent preemption and reschedule on another processor */ > > > > + int me = get_cpu(); > > > > + > > > > + /* Can deadlock when called with interrupts disabled */ > > > > + WARN_ON(wait && irqs_disabled()); > > > > > > With this fallback to wait the above condition isn't sufficient. > > > > What deadlock are you concerned with here? Would making cfd_fallback > > per-cpu make you feel better? > > CPU0 CPU1 > > local_irq_disable() local_irq_disable() > > smp_call_function_single(0,..,0) > test_and_set_bit_lock() > send IPI > smp_call_function_single(1,..,0) > while(test_and_set_bit_lock()) > cpu_relax(); > > > This will spin forever, because it needs to handle the IPI in order to > free the cfd_fallback thingy, but can't for its waiting for it. > > That particular deadlock can indeed be solved by making cfd_fallback > per-cpu. Right, that is the case I was thinking of. I added per-cpu fallbacks to cover that case. > But if you were to use multiple smp_call_function*() calls under a > single IRQ disabled, then that would not be sufficient. Now I can't > directly come up with a good reason to need to do that, but still. > > You'd need somethine like: > > local_irq_disable() > > smp_call_function_single(n, func_a,..,0) > smp_call_function_single(m, func_b,..,0) > > local_irq_enable() > > And invite 3 cpus to the party while under memory pressure and you get > deadlock potential. > > [ if it were both the same function, you'd want to use > smp_call_function() and provide a mask; if it were the same cpu you'd > want to call a function doing both ] I think that is plenty far off into theoretical country that we can get by with just documenting this limitation. Nobody is doing that currently in the kernel, and I see no practical use case for it. -- Jens Axboe From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from brick.kernel.dk ([87.55.233.238]:8535 "EHLO kernel.dk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751530AbYDWHti (ORCPT ); Wed, 23 Apr 2008 03:49:38 -0400 Date: Wed, 23 Apr 2008 09:49:33 +0200 From: Jens Axboe Subject: Re: [PATCH 1/11] Add generic helpers for arch IPI function calls Message-ID: <20080423074933.GB12774@kernel.dk> References: <1208890227-24808-1-git-send-email-jens.axboe@oracle.com> <1208890227-24808-2-git-send-email-jens.axboe@oracle.com> <1208895423.7115.290.camel@twins> <20080423060725.GT12774@kernel.dk> <1208932320.7115.319.camel@twins> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1208932320.7115.319.camel@twins> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Peter Zijlstra Cc: linux-arch@vger.kernel.org, linux-kernel@vger.kernel.org, npiggin@suse.de, torvalds@linux-foundation.org, sam@ravnborg.org Message-ID: <20080423074933.PxKukr6udaz00ICcc5S3WCQ6pFggVSh-gc38gIvgCpU@z> On Wed, Apr 23 2008, Peter Zijlstra wrote: > On Wed, 2008-04-23 at 08:07 +0200, Jens Axboe wrote: > > On Tue, Apr 22 2008, Peter Zijlstra wrote: > > > On Tue, 2008-04-22 at 20:50 +0200, Jens Axboe wrote: > > > > > > > +int smp_call_function_single(int cpu, void (*func) (void *info), void *info, > > > > + int retry, int wait) > > > > +{ > > > > + unsigned long flags; > > > > + /* prevent preemption and reschedule on another processor */ > > > > + int me = get_cpu(); > > > > + > > > > + /* Can deadlock when called with interrupts disabled */ > > > > + WARN_ON(wait && irqs_disabled()); > > > > > > With this fallback to wait the above condition isn't sufficient. > > > > What deadlock are you concerned with here? Would making cfd_fallback > > per-cpu make you feel better? > > CPU0 CPU1 > > local_irq_disable() local_irq_disable() > > smp_call_function_single(0,..,0) > test_and_set_bit_lock() > send IPI > smp_call_function_single(1,..,0) > while(test_and_set_bit_lock()) > cpu_relax(); > > > This will spin forever, because it needs to handle the IPI in order to > free the cfd_fallback thingy, but can't for its waiting for it. > > That particular deadlock can indeed be solved by making cfd_fallback > per-cpu. Right, that is the case I was thinking of. I added per-cpu fallbacks to cover that case. > But if you were to use multiple smp_call_function*() calls under a > single IRQ disabled, then that would not be sufficient. Now I can't > directly come up with a good reason to need to do that, but still. > > You'd need somethine like: > > local_irq_disable() > > smp_call_function_single(n, func_a,..,0) > smp_call_function_single(m, func_b,..,0) > > local_irq_enable() > > And invite 3 cpus to the party while under memory pressure and you get > deadlock potential. > > [ if it were both the same function, you'd want to use > smp_call_function() and provide a mask; if it were the same cpu you'd > want to call a function doing both ] I think that is plenty far off into theoretical country that we can get by with just documenting this limitation. Nobody is doing that currently in the kernel, and I see no practical use case for it. -- Jens Axboe