From mboxrd@z Thu Jan 1 00:00:00 1970 From: Avi Kivity Subject: Re: [PATCH v6 08/12] Handle async PF in a guest. Date: Thu, 07 Oct 2010 19:18:03 +0200 Message-ID: <4CAE00CB.1070400@redhat.com> References: <1286207794-16120-1-git-send-email-gleb@redhat.com> <1286207794-16120-9-git-send-email-gleb@redhat.com> <4CADC6C3.3040305@redhat.com> <20101007171418.GA2397@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: kvm@vger.kernel.org, linux-mm@kvack.org, linux-kernel@vger.kernel.org, mingo@elte.hu, a.p.zijlstra@chello.nl, tglx@linutronix.de, hpa@zytor.com, riel@redhat.com, cl@linux-foundation.org, mtosatti@redhat.com To: Gleb Natapov Return-path: Received: from mx1.redhat.com ([209.132.183.28]:65237 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751514Ab0JGRSc (ORCPT ); Thu, 7 Oct 2010 13:18:32 -0400 In-Reply-To: <20101007171418.GA2397@redhat.com> Sender: kvm-owner@vger.kernel.org List-ID: On 10/07/2010 07:14 PM, Gleb Natapov wrote: > On Thu, Oct 07, 2010 at 03:10:27PM +0200, Avi Kivity wrote: > > On 10/04/2010 05:56 PM, Gleb Natapov wrote: > > >When async PF capability is detected hook up special page fault handler > > >that will handle async page fault events and bypass other page faults to > > >regular page fault handler. Also add async PF handling to nested SVM > > >emulation. Async PF always generates exit to L1 where vcpu thread will > > >be scheduled out until page is available. > > > > > > > Please separate guest and host changes. > > > > >+void kvm_async_pf_task_wait(u32 token) > > >+{ > > >+ u32 key = hash_32(token, KVM_TASK_SLEEP_HASHBITS); > > >+ struct kvm_task_sleep_head *b =&async_pf_sleepers[key]; > > >+ struct kvm_task_sleep_node n, *e; > > >+ DEFINE_WAIT(wait); > > >+ > > >+ spin_lock(&b->lock); > > >+ e = _find_apf_task(b, token); > > >+ if (e) { > > >+ /* dummy entry exist -> wake up was delivered ahead of PF */ > > >+ hlist_del(&e->link); > > >+ kfree(e); > > >+ spin_unlock(&b->lock); > > >+ return; > > >+ } > > >+ > > >+ n.token = token; > > >+ n.cpu = smp_processor_id(); > > >+ init_waitqueue_head(&n.wq); > > >+ hlist_add_head(&n.link,&b->list); > > >+ spin_unlock(&b->lock); > > >+ > > >+ for (;;) { > > >+ prepare_to_wait(&n.wq,&wait, TASK_UNINTERRUPTIBLE); > > >+ if (hlist_unhashed(&n.link)) > > >+ break; > > >+ local_irq_enable(); > > > > Suppose we take another apf here. And another, and another (for > > different pages, while executing schedule()). What's to prevent > > kernel stack overflow? > > > Host side keeps track of outstanding apfs and will not send apf for the > same phys address twice. It will halt vcpu instead. What about different pages, running the scheduler code? Oh, and we'll run the scheduler recursively. -- I have a truly marvellous patch that fixes the bug which this signature is too narrow to contain.