[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] [PATCH V3 4/6] virtio-pci: use zero length mmio eventfd for
From: |
Jason Wang |
Subject: |
[Qemu-devel] [PATCH V3 4/6] virtio-pci: use zero length mmio eventfd for 1.0 notification cap when possible |
Date: |
Fri, 6 Nov 2015 16:02:47 +0800 |
We use data match eventfd for 1.0 notification currently. This could
be slow since software decoding is needed for mmio exit. To speed this
up, we can switch to use zero length mmio eventfd for 1.0 notification
since we can examine the queue index directly from the writing
address. KVM kernel module can utilize this by registering it to fast
mmio bus which could be as fast as pio on ept capable machine when
fast mmio is supported by host kernel.
Lots of improvements were seen on a ept capable machine:
Guest RX:(TCP)
size/session/+throughput%/+cpu%/-+per cpu%/
64/1/+1.6807%/[-16.2421%]/[+21.3984%]/
64/2/+0.6091%/[-11.0187%]/[+13.0678%]/
64/4/+0.0553%/[-5.9768%]/[+6.4155%]/
64/8/+0.1206%/[-4.0057%]/[+4.2984%]/
256/1/-0.0031%/[-10.1166%]/[+11.2517%]/
256/2/-0.5058%/[-6.1656%]/+6.0317%]/
...
Guest TX:(TCP)
size/session/+throughput%/+cpu%/-+per cpu%/
64/1/[+18.9183%]/-0.2823%/[+19.2550%]/
64/2/[+13.5714%]/[+2.2675%]/[+11.0533%]/
64/4/[+13.1070%]/[+2.1817%]/[+10.6920%]/
64/8/[+13.0426%]/[+2.0887%]/[+10.7299%]/
256/1/[+36.2761%]/+6.3434%/[+28.1471%]/
...
1024/1/[+44.8873%]/+2.0811%/[+41.9335%]/
...
1024/4/+0.0228%/[-2.2044%]/[+2.2774%]/
...
16384/2/+0.0127%/[-5.0346%]/[+5.3148%]/
...
65535/1/[+0.0062%]/[-4.1183%]/[+4.3017%]/
65535/2/+0.0004%/[-4.2311%]/[+4.4185%]/
65535/4/+0.0107%/[-4.6106%]/[+4.8446%]/
65535/8/-0.0090%/[-5.5178%]/[+5.8306%]/
Latency:(TCP_RR)
size/session/+transaction rate%/+cpu%/-+per cpu%/
64/1/[+6.5248%]/[-9.2882%]/[+17.4322%]/
64/25/[+11.0854%]/[+0.8000%]/[+10.2038%]/
64/50/[+12.1076%]/[+2.4627%]/[+9.4131%]/
256/1/[+5.3677%]/[+10.5669%]/-4.7024%/
256/25/[+5.6402%]/-0.8962%/[+6.5955%]/
256/50/[+5.9685%]/[+1.7766%]/[+4.1188%]/
4096/1/+0.2508%/[-10.4941%]/[+12.0047%]/
4096/25/[+1.8533%]/-0.0273%/+1.8812%/
4096/50/[+1.2156%]/-1.4134%/+2.6667%/
Notes: data with '[]' is the one whose significance is greater than 95%.
Thanks Wenli Quan <address@hidden> for the benchmarking.
Cc: Michael S. Tsirkin <address@hidden>
Signed-off-by: Jason Wang <address@hidden>
---
hw/virtio/virtio-pci.c | 19 +++++++++++++++----
1 file changed, 15 insertions(+), 4 deletions(-)
diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c
index ab2a372..9c3f6e3 100644
--- a/hw/virtio/virtio-pci.c
+++ b/hw/virtio/virtio-pci.c
@@ -270,6 +270,7 @@ static int
virtio_pci_set_host_notifier_internal(VirtIOPCIProxy *proxy,
EventNotifier *notifier = virtio_queue_get_host_notifier(vq);
bool legacy = !(proxy->flags & VIRTIO_PCI_FLAG_DISABLE_LEGACY);
bool modern = !(proxy->flags & VIRTIO_PCI_FLAG_DISABLE_MODERN);
+ bool fast_mmio = kvm_ioeventfd_any_length_enabled();
MemoryRegion *modern_mr = &proxy->notify.mr;
MemoryRegion *legacy_mr = &proxy->bar;
hwaddr modern_addr = QEMU_VIRTIO_PCI_QUEUE_MEM_MULT *
@@ -286,8 +287,13 @@ static int
virtio_pci_set_host_notifier_internal(VirtIOPCIProxy *proxy,
}
virtio_queue_set_host_notifier_fd_handler(vq, true, set_handler);
if (modern) {
- memory_region_add_eventfd(modern_mr, modern_addr, 2,
- true, n, notifier);
+ if (fast_mmio) {
+ memory_region_add_eventfd(modern_mr, modern_addr, 0,
+ false, n, notifier);
+ } else {
+ memory_region_add_eventfd(modern_mr, modern_addr, 2,
+ false, n, notifier);
+ }
}
if (legacy) {
memory_region_add_eventfd(legacy_mr, legacy_addr, 2,
@@ -295,8 +301,13 @@ static int
virtio_pci_set_host_notifier_internal(VirtIOPCIProxy *proxy,
}
} else {
if (modern) {
- memory_region_del_eventfd(modern_mr, modern_addr, 2,
- true, n, notifier);
+ if (fast_mmio) {
+ memory_region_del_eventfd(modern_mr, modern_addr, 0,
+ false, n, notifier);
+ } else {
+ memory_region_del_eventfd(modern_mr, modern_addr, 2,
+ false, n, notifier);
+ }
}
if (legacy) {
memory_region_del_eventfd(legacy_mr, legacy_addr, 2,
--
2.1.4
- [Qemu-devel] [PATCH V3 0/6] virtio 1.0 pci optimizations and fixes, Jason Wang, 2015/11/06
- [Qemu-devel] [PATCH V3 1/6] virtio-pci: fix 1.0 virtqueue migration, Jason Wang, 2015/11/06
- [Qemu-devel] [PATCH V3 2/6] memory: don't try to adjust endianness for zero length eventfd, Jason Wang, 2015/11/06
- [Qemu-devel] [PATCH V3 3/6] KVM: add support for any length io eventfd, Jason Wang, 2015/11/06
- [Qemu-devel] [PATCH V3 4/6] virtio-pci: use zero length mmio eventfd for 1.0 notification cap when possible,
Jason Wang <=
- [Qemu-devel] [PATCH V3 5/6] virtio-pci: introduce pio notification capability for modern device, Jason Wang, 2015/11/06
- [Qemu-devel] [PATCH V3 6/6] virtio-pci: unbreak queue_enable read, Jason Wang, 2015/11/06