Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753615AbZJUOfG (ORCPT ); Wed, 21 Oct 2009 10:35:06 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753599AbZJUOfE (ORCPT ); Wed, 21 Oct 2009 10:35:04 -0400 Received: from victor.provo.novell.com ([137.65.250.26]:60073 "EHLO victor.provo.novell.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753520AbZJUOfC (ORCPT ); Wed, 21 Oct 2009 10:35:02 -0400 From: Gregory Haskins Subject: [KVM PATCH 1/2] KVM: Directly inject interrupts via irqfd To: kvm@vger.kernel.org Cc: linux-kernel@vger.kernel.org, alacrityvm-devel@lists.sourceforge.net Date: Wed, 21 Oct 2009 10:34:53 -0400 Message-ID: <20091021143453.14955.80578.stgit@dev.haskins.net> In-Reply-To: <20091021143042.14955.22470.stgit@dev.haskins.net> References: <20091021143042.14955.22470.stgit@dev.haskins.net> User-Agent: StGIT/0.14.3 MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2813 Lines: 83 IRQFD currently uses a deferred workqueue item to execute the injection operation. It was originally designed this way because kvm_set_irq() required the caller to hold the irq_lock mutex, and the eventfd callback is invoked from within a non-preemptible critical section. With the advent of lockless injection support in kvm_set_irq, the deferment mechanism is no longer technically needed. Since context switching to the workqueue is a source of interrupt latency, lets switch to a direct method. Signed-off-by: Gregory Haskins --- virt/kvm/eventfd.c | 15 +++------------ 1 files changed, 3 insertions(+), 12 deletions(-) diff --git a/virt/kvm/eventfd.c b/virt/kvm/eventfd.c index 30f70fd..1a529d4 100644 --- a/virt/kvm/eventfd.c +++ b/virt/kvm/eventfd.c @@ -49,16 +49,14 @@ struct _irqfd { poll_table pt; wait_queue_head_t *wqh; wait_queue_t wait; - struct work_struct inject; struct work_struct shutdown; }; static struct workqueue_struct *irqfd_cleanup_wq; static void -irqfd_inject(struct work_struct *work) +irqfd_inject(struct _irqfd *irqfd) { - struct _irqfd *irqfd = container_of(work, struct _irqfd, inject); struct kvm *kvm = irqfd->kvm; kvm_set_irq(kvm, KVM_USERSPACE_IRQ_SOURCE_ID, irqfd->gsi, 1); @@ -80,12 +78,6 @@ irqfd_shutdown(struct work_struct *work) remove_wait_queue(irqfd->wqh, &irqfd->wait); /* - * We know no new events will be scheduled at this point, so block - * until all previously outstanding events have completed - */ - flush_work(&irqfd->inject); - - /* * It is now safe to release the object's resources */ eventfd_ctx_put(irqfd->eventfd); @@ -126,7 +118,7 @@ irqfd_wakeup(wait_queue_t *wait, unsigned mode, int sync, void *key) if (flags & POLLIN) /* An event has been signaled, inject an interrupt */ - schedule_work(&irqfd->inject); + irqfd_inject(irqfd); if (flags & POLLHUP) { /* The eventfd is closing, detach from KVM */ @@ -179,7 +171,6 @@ kvm_irqfd_assign(struct kvm *kvm, int fd, int gsi) irqfd->kvm = kvm; irqfd->gsi = gsi; INIT_LIST_HEAD(&irqfd->list); - INIT_WORK(&irqfd->inject, irqfd_inject); INIT_WORK(&irqfd->shutdown, irqfd_shutdown); file = eventfd_fget(fd); @@ -214,7 +205,7 @@ kvm_irqfd_assign(struct kvm *kvm, int fd, int gsi) * before we registered, and trigger it as if we didn't miss it. */ if (events & POLLIN) - schedule_work(&irqfd->inject); + irqfd_inject(irqfd); /* * do not drop the file until the irqfd is fully initialized, otherwise -- 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/