From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933781AbaGOVfQ (ORCPT ); Tue, 15 Jul 2014 17:35:16 -0400 Received: from youngberry.canonical.com ([91.189.89.112]:54311 "EHLO youngberry.canonical.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933759AbaGOVfK (ORCPT ); Tue, 15 Jul 2014 17:35:10 -0400 From: Kamal Mostafa To: linux-kernel@vger.kernel.org, stable@vger.kernel.org, kernel-team@lists.ubuntu.com Cc: Rusty Russell , "Michael S. Tsirkin" , virtualization@lists.linux-foundation.org, Ming Lei , Jens Axboe , Kamal Mostafa Subject: [PATCH 3.13 077/198] block: virtio_blk: don't hold spin lock during world switch Date: Tue, 15 Jul 2014 14:29:07 -0700 Message-Id: <1405459868-15089-78-git-send-email-kamal@canonical.com> X-Mailer: git-send-email 1.9.1 In-Reply-To: <1405459868-15089-1-git-send-email-kamal@canonical.com> References: <1405459868-15089-1-git-send-email-kamal@canonical.com> X-Extended-Stable: 3.13 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org 3.13.11.5 -stable review patch. If anyone has any objections, please let me know. ------------------ From: Ming Lei commit e8edca6f7f92234202d6dd163c118ef495244d7c upstream. Firstly, it isn't necessary to hold lock of vblk->vq_lock when notifying hypervisor about queued I/O. Secondly, virtqueue_notify() will cause world switch and it may take long time on some hypervisors(such as, qemu-arm), so it isn't good to hold the lock and block other vCPUs. On arm64 quad core VM(qemu-kvm), the patch can increase I/O performance a lot with VIRTIO_RING_F_EVENT_IDX enabled: - without the patch: 14K IOPS - with the patch: 34K IOPS fio script: [global] direct=1 bsrange=4k-4k timeout=10 numjobs=4 ioengine=libaio iodepth=64 filename=/dev/vdc group_reporting=1 [f1] rw=randread Cc: Rusty Russell Cc: "Michael S. Tsirkin" Cc: virtualization@lists.linux-foundation.org Signed-off-by: Ming Lei Acked-by: Rusty Russell Signed-off-by: Jens Axboe Signed-off-by: Kamal Mostafa --- drivers/block/virtio_blk.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c index 9b734f2..53e3db6 100644 --- a/drivers/block/virtio_blk.c +++ b/drivers/block/virtio_blk.c @@ -159,6 +159,7 @@ static int virtio_queue_rq(struct blk_mq_hw_ctx *hctx, struct request *req) unsigned int num; const bool last = (req->cmd_flags & REQ_END) != 0; int err; + bool notify = false; BUG_ON(req->nr_phys_segments + 2 > vblk->sg_elems); @@ -211,10 +212,12 @@ static int virtio_queue_rq(struct blk_mq_hw_ctx *hctx, struct request *req) return BLK_MQ_RQ_QUEUE_ERROR; } - if (last) - virtqueue_kick(vblk->vq); - + if (last && virtqueue_kick_prepare(vblk->vq)) + notify = true; spin_unlock_irqrestore(&vblk->vq_lock, flags); + + if (notify) + virtqueue_notify(vblk->vq); return BLK_MQ_RQ_QUEUE_OK; } -- 1.9.1