From 0706d1c4cad581d9f213e742a518be939b05a733 Mon Sep 17 00:00:00 2001
From: Stefan Hajnoczi <stefanha@redhat.com>
Date: Fri, 22 Apr 2016 10:09:35 +0200
Subject: [PATCH 5/5] virtio-blk: use blk_io_plug/unplug for Linux AIO batching
RH-Author: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: <1461319775-24446-2-git-send-email-stefanha@redhat.com>
Patchwork-id: 70224
O-Subject: [RHEV-7.2.z qemu-kvm-rhev PATCH 1/1] virtio-blk: use blk_io_plug/unplug for Linux AIO batching
Bugzilla: 1327613
RH-Acked-by: Laurent Vivier <lvivier@redhat.com>
RH-Acked-by: Fam Zheng <famz@redhat.com>
RH-Acked-by: Kevin Wolf <kwolf@redhat.com>
The raw-posix block driver implements Linux AIO batching so multiple
requests can be submitted with a single io_submit(2) system call.
Batching is currently only used by virtio-scsi and
virtio-blk-data-plane.
Enable batching for regular virtio-blk so the number of io_submit(2)
system calls is reduced for workloads with queue depth > 1.
In 4KB random read performance tests with queue depth 32, the CPU
utilization on the host is reduced by 9.4%. The fio job is as follows:
[global]
bs=4k
ioengine=libaio
iodepth=32
direct=1
sync=0
time_based=1
runtime=30
clocksource=gettimeofday
ramp_time=5
[job1]
rw=randread
filename=/dev/vdb
size=4096M
write_bw_log=fio
write_iops_log=fio
write_lat_log=fio
log_avg_msec=1000
This benchmark was run on an raw image on LVM. The disk was an SSD
drive and -drive cache=none,aio=native was used.
Tested-by: Pradeep Surisetty <psuriset@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
(cherry-picked from commit fc73548e444ae3239f6cef44a5200b5d2c3e85d1)
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Miroslav Rezanina <mrezanin@redhat.com>
---
hw/block/virtio-blk.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index 6dd7fed..c99aefd 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -604,6 +604,8 @@ static void virtio_blk_handle_output(VirtIODevice *vdev, VirtQueue *vq)
return;
}
+ blk_io_plug(s->blk);
+
while ((req = virtio_blk_get_request(s))) {
virtio_blk_handle_request(req, &mrb);
}
@@ -611,6 +613,8 @@ static void virtio_blk_handle_output(VirtIODevice *vdev, VirtQueue *vq)
if (mrb.num_reqs) {
virtio_blk_submit_multireq(s->blk, &mrb);
}
+
+ blk_io_unplug(s->blk);
}
static void virtio_blk_dma_restart_bh(void *opaque)
--
1.8.3.1