qemu/hw/virtio/virtio-pci.c

2065 lines
66 KiB
C
Raw Normal View History

/*
* Virtio PCI Bindings
*
* Copyright IBM, Corp. 2007
* Copyright (c) 2009 CodeSourcery
*
* Authors:
* Anthony Liguori <aliguori@us.ibm.com>
* Paul Brook <paul@codesourcery.com>
*
* This work is licensed under the terms of the GNU GPL, version 2. See
* the COPYING file in the top-level directory.
*
* Contributions after 2012-01-13 are licensed under the terms of the
* GNU GPL, version 2 or (at your option) any later version.
*/
#include "qemu/osdep.h"
#include "standard-headers/linux/virtio_pci.h"
#include "hw/virtio/virtio.h"
#include "hw/pci/pci.h"
#include "hw/pci/pci_bus.h"
2016-03-14 11:01:28 +03:00
#include "qapi/error.h"
#include "qemu/error-report.h"
#include "qemu/module.h"
#include "hw/pci/msi.h"
#include "hw/pci/msix.h"
#include "hw/loader.h"
#include "sysemu/kvm.h"
#include "virtio-pci.h"
#include "qemu/range.h"
#include "hw/virtio/virtio-bus.h"
#include "qapi/visitor.h"
#define VIRTIO_PCI_REGION_SIZE(dev) VIRTIO_PCI_CONFIG_OFF(msix_present(dev))
#undef VIRTIO_PCI_CONFIG
/* The remaining space is defined by each driver as the per-driver
* configuration space */
#define VIRTIO_PCI_CONFIG_SIZE(dev) VIRTIO_PCI_CONFIG_OFF(msix_enabled(dev))
static void virtio_pci_bus_new(VirtioBusState *bus, size_t bus_size,
VirtIOPCIProxy *dev);
static void virtio_pci_reset(DeviceState *qdev);
/* virtio device */
/* DeviceState to VirtIOPCIProxy. For use off data-path. TODO: use QOM. */
static inline VirtIOPCIProxy *to_virtio_pci_proxy(DeviceState *d)
{
return container_of(d, VirtIOPCIProxy, pci_dev.qdev);
}
/* DeviceState to VirtIOPCIProxy. Note: used on datapath,
* be careful and test performance if you change this.
*/
static inline VirtIOPCIProxy *to_virtio_pci_proxy_fast(DeviceState *d)
{
return container_of(d, VirtIOPCIProxy, pci_dev.qdev);
}
static void virtio_pci_notify(DeviceState *d, uint16_t vector)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy_fast(d);
if (msix_enabled(&proxy->pci_dev))
msix_notify(&proxy->pci_dev, vector);
else {
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
pci_set_irq(&proxy->pci_dev, atomic_read(&vdev->isr) & 1);
}
}
static void virtio_pci_save_config(DeviceState *d, QEMUFile *f)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
pci_device_save(&proxy->pci_dev, f);
msix_save(&proxy->pci_dev, f);
if (msix_present(&proxy->pci_dev))
qemu_put_be16(f, vdev->config_vector);
}
static const VMStateDescription vmstate_virtio_pci_modern_queue_state = {
.name = "virtio_pci/modern_queue_state",
.version_id = 1,
.minimum_version_id = 1,
.fields = (VMStateField[]) {
VMSTATE_UINT16(num, VirtIOPCIQueue),
VMSTATE_UNUSED(1), /* enabled was stored as be16 */
VMSTATE_BOOL(enabled, VirtIOPCIQueue),
VMSTATE_UINT32_ARRAY(desc, VirtIOPCIQueue, 2),
VMSTATE_UINT32_ARRAY(avail, VirtIOPCIQueue, 2),
VMSTATE_UINT32_ARRAY(used, VirtIOPCIQueue, 2),
VMSTATE_END_OF_LIST()
}
};
static bool virtio_pci_modern_state_needed(void *opaque)
{
VirtIOPCIProxy *proxy = opaque;
return virtio_pci_modern(proxy);
}
static const VMStateDescription vmstate_virtio_pci_modern_state_sub = {
.name = "virtio_pci/modern_state",
.version_id = 1,
.minimum_version_id = 1,
.needed = &virtio_pci_modern_state_needed,
.fields = (VMStateField[]) {
VMSTATE_UINT32(dfselect, VirtIOPCIProxy),
VMSTATE_UINT32(gfselect, VirtIOPCIProxy),
VMSTATE_UINT32_ARRAY(guest_features, VirtIOPCIProxy, 2),
VMSTATE_STRUCT_ARRAY(vqs, VirtIOPCIProxy, VIRTIO_QUEUE_MAX, 0,
vmstate_virtio_pci_modern_queue_state,
VirtIOPCIQueue),
VMSTATE_END_OF_LIST()
}
};
static const VMStateDescription vmstate_virtio_pci = {
.name = "virtio_pci",
.version_id = 1,
.minimum_version_id = 1,
.minimum_version_id_old = 1,
.fields = (VMStateField[]) {
VMSTATE_END_OF_LIST()
},
.subsections = (const VMStateDescription*[]) {
&vmstate_virtio_pci_modern_state_sub,
NULL
}
};
static bool virtio_pci_has_extra_state(DeviceState *d)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
return proxy->flags & VIRTIO_PCI_FLAG_MIGRATE_EXTRA;
}
static void virtio_pci_save_extra_state(DeviceState *d, QEMUFile *f)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
vmstate_save_state(f, &vmstate_virtio_pci, proxy, NULL);
}
static int virtio_pci_load_extra_state(DeviceState *d, QEMUFile *f)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
return vmstate_load_state(f, &vmstate_virtio_pci, proxy, 1);
}
static void virtio_pci_save_queue(DeviceState *d, int n, QEMUFile *f)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
if (msix_present(&proxy->pci_dev))
qemu_put_be16(f, virtio_queue_vector(vdev, n));
}
static int virtio_pci_load_config(DeviceState *d, QEMUFile *f)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
int ret;
ret = pci_device_load(&proxy->pci_dev, f);
if (ret) {
return ret;
}
msix_unuse_all_vectors(&proxy->pci_dev);
msix_load(&proxy->pci_dev, f);
if (msix_present(&proxy->pci_dev)) {
qemu_get_be16s(f, &vdev->config_vector);
} else {
vdev->config_vector = VIRTIO_NO_VECTOR;
}
if (vdev->config_vector != VIRTIO_NO_VECTOR) {
return msix_vector_use(&proxy->pci_dev, vdev->config_vector);
}
return 0;
}
static int virtio_pci_load_queue(DeviceState *d, int n, QEMUFile *f)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
uint16_t vector;
if (msix_present(&proxy->pci_dev)) {
qemu_get_be16s(f, &vector);
} else {
vector = VIRTIO_NO_VECTOR;
}
virtio_queue_set_vector(vdev, n, vector);
if (vector != VIRTIO_NO_VECTOR) {
return msix_vector_use(&proxy->pci_dev, vector);
}
return 0;
}
static bool virtio_pci_ioeventfd_enabled(DeviceState *d)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
return (proxy->flags & VIRTIO_PCI_FLAG_USE_IOEVENTFD) != 0;
}
#define QEMU_VIRTIO_PCI_QUEUE_MEM_MULT 0x1000
static inline int virtio_pci_queue_mem_mult(struct VirtIOPCIProxy *proxy)
{
return (proxy->flags & VIRTIO_PCI_FLAG_PAGE_PER_VQ) ?
QEMU_VIRTIO_PCI_QUEUE_MEM_MULT : 4;
}
static int virtio_pci_ioeventfd_assign(DeviceState *d, EventNotifier *notifier,
int n, bool assign)
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
VirtQueue *vq = virtio_get_queue(vdev, n);
bool legacy = virtio_pci_legacy(proxy);
bool modern = virtio_pci_modern(proxy);
virtio-pci: use zero length mmio eventfd for 1.0 notification cap when possible 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 <wquan@redhat.com> for the benchmarking. Cc: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Jason Wang <jasowang@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2015-11-06 11:02:47 +03:00
bool fast_mmio = kvm_ioeventfd_any_length_enabled();
bool modern_pio = proxy->flags & VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY;
MemoryRegion *modern_mr = &proxy->notify.mr;
MemoryRegion *modern_notify_mr = &proxy->notify_pio.mr;
MemoryRegion *legacy_mr = &proxy->bar;
hwaddr modern_addr = virtio_pci_queue_mem_mult(proxy) *
virtio_get_queue_index(vq);
hwaddr legacy_addr = VIRTIO_PCI_QUEUE_NOTIFY;
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
if (assign) {
if (modern) {
virtio-pci: use zero length mmio eventfd for 1.0 notification cap when possible 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 <wquan@redhat.com> for the benchmarking. Cc: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Jason Wang <jasowang@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2015-11-06 11:02:47 +03:00
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 (modern_pio) {
memory_region_add_eventfd(modern_notify_mr, 0, 2,
true, n, notifier);
}
}
if (legacy) {
memory_region_add_eventfd(legacy_mr, legacy_addr, 2,
true, n, notifier);
}
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
} else {
if (modern) {
virtio-pci: use zero length mmio eventfd for 1.0 notification cap when possible 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 <wquan@redhat.com> for the benchmarking. Cc: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Jason Wang <jasowang@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2015-11-06 11:02:47 +03:00
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 (modern_pio) {
memory_region_del_eventfd(modern_notify_mr, 0, 2,
true, n, notifier);
}
}
if (legacy) {
memory_region_del_eventfd(legacy_mr, legacy_addr, 2,
true, n, notifier);
}
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
}
return 0;
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
}
static void virtio_pci_start_ioeventfd(VirtIOPCIProxy *proxy)
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
{
virtio_bus_start_ioeventfd(&proxy->bus);
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
}
static void virtio_pci_stop_ioeventfd(VirtIOPCIProxy *proxy)
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
{
virtio_bus_stop_ioeventfd(&proxy->bus);
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
}
static void virtio_ioport_write(void *opaque, uint32_t addr, uint32_t val)
{
VirtIOPCIProxy *proxy = opaque;
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
hwaddr pa;
switch (addr) {
case VIRTIO_PCI_GUEST_FEATURES:
/* Guest does not negotiate properly? We have to assume nothing. */
if (val & (1 << VIRTIO_F_BAD_FEATURE)) {
val = virtio_bus_get_vdev_bad_features(&proxy->bus);
}
virtio_set_features(vdev, val);
break;
case VIRTIO_PCI_QUEUE_PFN:
pa = (hwaddr)val << VIRTIO_PCI_QUEUE_ADDR_SHIFT;
if (pa == 0) {
virtio_pci_reset(DEVICE(proxy));
}
else
virtio_queue_set_addr(vdev, vdev->queue_sel, pa);
break;
case VIRTIO_PCI_QUEUE_SEL:
if (val < VIRTIO_QUEUE_MAX)
vdev->queue_sel = val;
break;
case VIRTIO_PCI_QUEUE_NOTIFY:
if (val < VIRTIO_QUEUE_MAX) {
virtio_queue_notify(vdev, val);
}
break;
case VIRTIO_PCI_STATUS:
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
if (!(val & VIRTIO_CONFIG_S_DRIVER_OK)) {
virtio_pci_stop_ioeventfd(proxy);
}
virtio_set_status(vdev, val & 0xFF);
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
if (val & VIRTIO_CONFIG_S_DRIVER_OK) {
virtio_pci_start_ioeventfd(proxy);
}
if (vdev->status == 0) {
virtio_pci_reset(DEVICE(proxy));
}
/* Linux before 2.6.34 drives the device without enabling
the PCI device bus master bit. Enable it automatically
for the guest. This is a PCI spec violation but so is
initiating DMA with bus master bit clear. */
if (val == (VIRTIO_CONFIG_S_ACKNOWLEDGE | VIRTIO_CONFIG_S_DRIVER)) {
pci_default_write_config(&proxy->pci_dev, PCI_COMMAND,
proxy->pci_dev.config[PCI_COMMAND] |
PCI_COMMAND_MASTER, 1);
}
break;
case VIRTIO_MSI_CONFIG_VECTOR:
msix_vector_unuse(&proxy->pci_dev, vdev->config_vector);
/* Make it possible for guest to discover an error took place. */
if (msix_vector_use(&proxy->pci_dev, val) < 0)
val = VIRTIO_NO_VECTOR;
vdev->config_vector = val;
break;
case VIRTIO_MSI_QUEUE_VECTOR:
msix_vector_unuse(&proxy->pci_dev,
virtio_queue_vector(vdev, vdev->queue_sel));
/* Make it possible for guest to discover an error took place. */
if (msix_vector_use(&proxy->pci_dev, val) < 0)
val = VIRTIO_NO_VECTOR;
virtio_queue_set_vector(vdev, vdev->queue_sel, val);
break;
default:
error_report("%s: unexpected address 0x%x value 0x%x",
__func__, addr, val);
break;
}
}
static uint32_t virtio_ioport_read(VirtIOPCIProxy *proxy, uint32_t addr)
{
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
uint32_t ret = 0xFFFFFFFF;
switch (addr) {
case VIRTIO_PCI_HOST_FEATURES:
ret = vdev->host_features;
break;
case VIRTIO_PCI_GUEST_FEATURES:
ret = vdev->guest_features;
break;
case VIRTIO_PCI_QUEUE_PFN:
ret = virtio_queue_get_addr(vdev, vdev->queue_sel)
>> VIRTIO_PCI_QUEUE_ADDR_SHIFT;
break;
case VIRTIO_PCI_QUEUE_NUM:
ret = virtio_queue_get_num(vdev, vdev->queue_sel);
break;
case VIRTIO_PCI_QUEUE_SEL:
ret = vdev->queue_sel;
break;
case VIRTIO_PCI_STATUS:
ret = vdev->status;
break;
case VIRTIO_PCI_ISR:
/* reading from the ISR also clears it. */
ret = atomic_xchg(&vdev->isr, 0);
pci_irq_deassert(&proxy->pci_dev);
break;
case VIRTIO_MSI_CONFIG_VECTOR:
ret = vdev->config_vector;
break;
case VIRTIO_MSI_QUEUE_VECTOR:
ret = virtio_queue_vector(vdev, vdev->queue_sel);
break;
default:
break;
}
return ret;
}
static uint64_t virtio_pci_config_read(void *opaque, hwaddr addr,
unsigned size)
{
VirtIOPCIProxy *proxy = opaque;
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
uint32_t config = VIRTIO_PCI_CONFIG_SIZE(&proxy->pci_dev);
uint64_t val = 0;
if (addr < config) {
return virtio_ioport_read(proxy, addr);
}
addr -= config;
switch (size) {
case 1:
val = virtio_config_readb(vdev, addr);
break;
case 2:
val = virtio_config_readw(vdev, addr);
if (virtio_is_big_endian(vdev)) {
val = bswap16(val);
}
break;
case 4:
val = virtio_config_readl(vdev, addr);
if (virtio_is_big_endian(vdev)) {
val = bswap32(val);
}
break;
}
return val;
}
static void virtio_pci_config_write(void *opaque, hwaddr addr,
uint64_t val, unsigned size)
{
VirtIOPCIProxy *proxy = opaque;
uint32_t config = VIRTIO_PCI_CONFIG_SIZE(&proxy->pci_dev);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
if (addr < config) {
virtio_ioport_write(proxy, addr, val);
return;
}
addr -= config;
/*
* Virtio-PCI is odd. Ioports are LE but config space is target native
* endian.
*/
switch (size) {
case 1:
virtio_config_writeb(vdev, addr, val);
break;
case 2:
if (virtio_is_big_endian(vdev)) {
val = bswap16(val);
}
virtio_config_writew(vdev, addr, val);
break;
case 4:
if (virtio_is_big_endian(vdev)) {
val = bswap32(val);
}
virtio_config_writel(vdev, addr, val);
break;
}
}
static const MemoryRegionOps virtio_pci_config_ops = {
.read = virtio_pci_config_read,
.write = virtio_pci_config_write,
.impl = {
.min_access_size = 1,
.max_access_size = 4,
},
.endianness = DEVICE_LITTLE_ENDIAN,
};
static MemoryRegion *virtio_address_space_lookup(VirtIOPCIProxy *proxy,
hwaddr *off, int len)
{
int i;
VirtIOPCIRegion *reg;
for (i = 0; i < ARRAY_SIZE(proxy->regs); ++i) {
reg = &proxy->regs[i];
if (*off >= reg->offset &&
*off + len <= reg->offset + reg->size) {
*off -= reg->offset;
return &reg->mr;
}
}
return NULL;
}
/* Below are generic functions to do memcpy from/to an address space,
* without byteswaps, with input validation.
*
* As regular address_space_* APIs all do some kind of byteswap at least for
* some host/target combinations, we are forced to explicitly convert to a
* known-endianness integer value.
* It doesn't really matter which endian format to go through, so the code
* below selects the endian that causes the least amount of work on the given
* host.
*
* Note: host pointer must be aligned.
*/
static
void virtio_address_space_write(VirtIOPCIProxy *proxy, hwaddr addr,
const uint8_t *buf, int len)
{
uint64_t val;
MemoryRegion *mr;
/* address_space_* APIs assume an aligned address.
* As address is under guest control, handle illegal values.
*/
addr &= ~(len - 1);
mr = virtio_address_space_lookup(proxy, &addr, len);
if (!mr) {
return;
}
/* Make sure caller aligned buf properly */
assert(!(((uintptr_t)buf) & (len - 1)));
switch (len) {
case 1:
val = pci_get_byte(buf);
break;
case 2:
val = cpu_to_le16(pci_get_word(buf));
break;
case 4:
val = cpu_to_le32(pci_get_long(buf));
break;
default:
/* As length is under guest control, handle illegal values. */
return;
}
memory_region_dispatch_write(mr, addr, val, len, MEMTXATTRS_UNSPECIFIED);
}
static void
virtio_address_space_read(VirtIOPCIProxy *proxy, hwaddr addr,
uint8_t *buf, int len)
{
uint64_t val;
MemoryRegion *mr;
/* address_space_* APIs assume an aligned address.
* As address is under guest control, handle illegal values.
*/
addr &= ~(len - 1);
mr = virtio_address_space_lookup(proxy, &addr, len);
if (!mr) {
return;
}
/* Make sure caller aligned buf properly */
assert(!(((uintptr_t)buf) & (len - 1)));
memory_region_dispatch_read(mr, addr, &val, len, MEMTXATTRS_UNSPECIFIED);
switch (len) {
case 1:
pci_set_byte(buf, val);
break;
case 2:
pci_set_word(buf, le16_to_cpu(val));
break;
case 4:
pci_set_long(buf, le32_to_cpu(val));
break;
default:
/* As length is under guest control, handle illegal values. */
break;
}
}
static void virtio_write_config(PCIDevice *pci_dev, uint32_t address,
uint32_t val, int len)
{
VirtIOPCIProxy *proxy = VIRTIO_PCI(pci_dev);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
struct virtio_pci_cfg_cap *cfg;
pci_default_write_config(pci_dev, address, val, len);
if (range_covers_byte(address, len, PCI_COMMAND) &&
!(pci_dev->config[PCI_COMMAND] & PCI_COMMAND_MASTER)) {
virtio_pci_stop_ioeventfd(proxy);
virtio_set_status(vdev, vdev->status & ~VIRTIO_CONFIG_S_DRIVER_OK);
}
if (proxy->config_cap &&
ranges_overlap(address, len, proxy->config_cap + offsetof(struct virtio_pci_cfg_cap,
pci_cfg_data),
sizeof cfg->pci_cfg_data)) {
uint32_t off;
uint32_t len;
cfg = (void *)(proxy->pci_dev.config + proxy->config_cap);
off = le32_to_cpu(cfg->cap.offset);
len = le32_to_cpu(cfg->cap.length);
if (len == 1 || len == 2 || len == 4) {
assert(len <= sizeof cfg->pci_cfg_data);
virtio_address_space_write(proxy, off, cfg->pci_cfg_data, len);
}
}
}
static uint32_t virtio_read_config(PCIDevice *pci_dev,
uint32_t address, int len)
{
VirtIOPCIProxy *proxy = VIRTIO_PCI(pci_dev);
struct virtio_pci_cfg_cap *cfg;
if (proxy->config_cap &&
ranges_overlap(address, len, proxy->config_cap + offsetof(struct virtio_pci_cfg_cap,
pci_cfg_data),
sizeof cfg->pci_cfg_data)) {
uint32_t off;
uint32_t len;
cfg = (void *)(proxy->pci_dev.config + proxy->config_cap);
off = le32_to_cpu(cfg->cap.offset);
len = le32_to_cpu(cfg->cap.length);
if (len == 1 || len == 2 || len == 4) {
assert(len <= sizeof cfg->pci_cfg_data);
virtio_address_space_read(proxy, off, cfg->pci_cfg_data, len);
}
}
return pci_default_read_config(pci_dev, address, len);
}
static int kvm_virtio_pci_vq_vector_use(VirtIOPCIProxy *proxy,
unsigned int queue_no,
unsigned int vector)
{
VirtIOIRQFD *irqfd = &proxy->vector_irqfd[vector];
int ret;
if (irqfd->users == 0) {
ret = kvm_irqchip_add_msi_route(kvm_state, vector, &proxy->pci_dev);
if (ret < 0) {
return ret;
}
irqfd->virq = ret;
}
irqfd->users++;
return 0;
}
static void kvm_virtio_pci_vq_vector_release(VirtIOPCIProxy *proxy,
unsigned int vector)
{
VirtIOIRQFD *irqfd = &proxy->vector_irqfd[vector];
if (--irqfd->users == 0) {
kvm_irqchip_release_virq(kvm_state, irqfd->virq);
}
}
static int kvm_virtio_pci_irqfd_use(VirtIOPCIProxy *proxy,
unsigned int queue_no,
unsigned int vector)
{
VirtIOIRQFD *irqfd = &proxy->vector_irqfd[vector];
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
VirtQueue *vq = virtio_get_queue(vdev, queue_no);
EventNotifier *n = virtio_queue_get_guest_notifier(vq);
return kvm_irqchip_add_irqfd_notifier_gsi(kvm_state, n, NULL, irqfd->virq);
}
static void kvm_virtio_pci_irqfd_release(VirtIOPCIProxy *proxy,
unsigned int queue_no,
unsigned int vector)
{
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
VirtQueue *vq = virtio_get_queue(vdev, queue_no);
EventNotifier *n = virtio_queue_get_guest_notifier(vq);
VirtIOIRQFD *irqfd = &proxy->vector_irqfd[vector];
int ret;
ret = kvm_irqchip_remove_irqfd_notifier_gsi(kvm_state, n, irqfd->virq);
assert(ret == 0);
}
static int kvm_virtio_pci_vector_use(VirtIOPCIProxy *proxy, int nvqs)
{
PCIDevice *dev = &proxy->pci_dev;
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
unsigned int vector;
int ret, queue_no;
for (queue_no = 0; queue_no < nvqs; queue_no++) {
if (!virtio_queue_get_num(vdev, queue_no)) {
break;
}
vector = virtio_queue_vector(vdev, queue_no);
if (vector >= msix_nr_vectors_allocated(dev)) {
continue;
}
ret = kvm_virtio_pci_vq_vector_use(proxy, queue_no, vector);
if (ret < 0) {
goto undo;
}
/* If guest supports masking, set up irqfd now.
* Otherwise, delay until unmasked in the frontend.
*/
if (vdev->use_guest_notifier_mask && k->guest_notifier_mask) {
ret = kvm_virtio_pci_irqfd_use(proxy, queue_no, vector);
if (ret < 0) {
kvm_virtio_pci_vq_vector_release(proxy, vector);
goto undo;
}
}
}
return 0;
undo:
while (--queue_no >= 0) {
vector = virtio_queue_vector(vdev, queue_no);
if (vector >= msix_nr_vectors_allocated(dev)) {
continue;
}
if (vdev->use_guest_notifier_mask && k->guest_notifier_mask) {
kvm_virtio_pci_irqfd_release(proxy, queue_no, vector);
}
kvm_virtio_pci_vq_vector_release(proxy, vector);
}
return ret;
}
static void kvm_virtio_pci_vector_release(VirtIOPCIProxy *proxy, int nvqs)
{
PCIDevice *dev = &proxy->pci_dev;
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
unsigned int vector;
int queue_no;
VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
for (queue_no = 0; queue_no < nvqs; queue_no++) {
if (!virtio_queue_get_num(vdev, queue_no)) {
break;
}
vector = virtio_queue_vector(vdev, queue_no);
if (vector >= msix_nr_vectors_allocated(dev)) {
continue;
}
/* If guest supports masking, clean up irqfd now.
* Otherwise, it was cleaned when masked in the frontend.
*/
if (vdev->use_guest_notifier_mask && k->guest_notifier_mask) {
kvm_virtio_pci_irqfd_release(proxy, queue_no, vector);
}
kvm_virtio_pci_vq_vector_release(proxy, vector);
}
}
static int virtio_pci_vq_vector_unmask(VirtIOPCIProxy *proxy,
unsigned int queue_no,
unsigned int vector,
MSIMessage msg)
{
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
VirtQueue *vq = virtio_get_queue(vdev, queue_no);
EventNotifier *n = virtio_queue_get_guest_notifier(vq);
VirtIOIRQFD *irqfd;
int ret = 0;
if (proxy->vector_irqfd) {
irqfd = &proxy->vector_irqfd[vector];
if (irqfd->msg.data != msg.data || irqfd->msg.address != msg.address) {
ret = kvm_irqchip_update_msi_route(kvm_state, irqfd->virq, msg,
&proxy->pci_dev);
if (ret < 0) {
return ret;
}
kvm_irqchip_commit_routes(kvm_state);
}
}
/* If guest supports masking, irqfd is already setup, unmask it.
* Otherwise, set it up now.
*/
if (vdev->use_guest_notifier_mask && k->guest_notifier_mask) {
k->guest_notifier_mask(vdev, queue_no, false);
/* Test after unmasking to avoid losing events. */
if (k->guest_notifier_pending &&
k->guest_notifier_pending(vdev, queue_no)) {
event_notifier_set(n);
}
} else {
ret = kvm_virtio_pci_irqfd_use(proxy, queue_no, vector);
}
return ret;
}
static void virtio_pci_vq_vector_mask(VirtIOPCIProxy *proxy,
unsigned int queue_no,
unsigned int vector)
{
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
/* If guest supports masking, keep irqfd but mask it.
* Otherwise, clean it up now.
*/
if (vdev->use_guest_notifier_mask && k->guest_notifier_mask) {
k->guest_notifier_mask(vdev, queue_no, true);
} else {
kvm_virtio_pci_irqfd_release(proxy, queue_no, vector);
}
}
static int virtio_pci_vector_unmask(PCIDevice *dev, unsigned vector,
MSIMessage msg)
{
VirtIOPCIProxy *proxy = container_of(dev, VirtIOPCIProxy, pci_dev);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
VirtQueue *vq = virtio_vector_first_queue(vdev, vector);
int ret, index, unmasked = 0;
while (vq) {
index = virtio_get_queue_index(vq);
if (!virtio_queue_get_num(vdev, index)) {
break;
}
if (index < proxy->nvqs_with_notifiers) {
ret = virtio_pci_vq_vector_unmask(proxy, index, vector, msg);
if (ret < 0) {
goto undo;
}
++unmasked;
}
vq = virtio_vector_next_queue(vq);
}
return 0;
undo:
vq = virtio_vector_first_queue(vdev, vector);
while (vq && unmasked >= 0) {
index = virtio_get_queue_index(vq);
if (index < proxy->nvqs_with_notifiers) {
virtio_pci_vq_vector_mask(proxy, index, vector);
--unmasked;
}
vq = virtio_vector_next_queue(vq);
}
return ret;
}
static void virtio_pci_vector_mask(PCIDevice *dev, unsigned vector)
{
VirtIOPCIProxy *proxy = container_of(dev, VirtIOPCIProxy, pci_dev);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
VirtQueue *vq = virtio_vector_first_queue(vdev, vector);
int index;
while (vq) {
index = virtio_get_queue_index(vq);
if (!virtio_queue_get_num(vdev, index)) {
break;
}
if (index < proxy->nvqs_with_notifiers) {
virtio_pci_vq_vector_mask(proxy, index, vector);
}
vq = virtio_vector_next_queue(vq);
}
}
static void virtio_pci_vector_poll(PCIDevice *dev,
unsigned int vector_start,
unsigned int vector_end)
{
VirtIOPCIProxy *proxy = container_of(dev, VirtIOPCIProxy, pci_dev);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
int queue_no;
unsigned int vector;
EventNotifier *notifier;
VirtQueue *vq;
for (queue_no = 0; queue_no < proxy->nvqs_with_notifiers; queue_no++) {
if (!virtio_queue_get_num(vdev, queue_no)) {
break;
}
vector = virtio_queue_vector(vdev, queue_no);
if (vector < vector_start || vector >= vector_end ||
!msix_is_masked(dev, vector)) {
continue;
}
vq = virtio_get_queue(vdev, queue_no);
notifier = virtio_queue_get_guest_notifier(vq);
if (k->guest_notifier_pending) {
if (k->guest_notifier_pending(vdev, queue_no)) {
msix_set_pending(dev, vector);
}
} else if (event_notifier_test_and_clear(notifier)) {
msix_set_pending(dev, vector);
}
}
}
static int virtio_pci_set_guest_notifier(DeviceState *d, int n, bool assign,
bool with_irqfd)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
VirtioDeviceClass *vdc = VIRTIO_DEVICE_GET_CLASS(vdev);
VirtQueue *vq = virtio_get_queue(vdev, n);
EventNotifier *notifier = virtio_queue_get_guest_notifier(vq);
if (assign) {
int r = event_notifier_init(notifier, 0);
if (r < 0) {
return r;
}
virtio_queue_set_guest_notifier_fd_handler(vq, true, with_irqfd);
} else {
virtio_queue_set_guest_notifier_fd_handler(vq, false, with_irqfd);
event_notifier_cleanup(notifier);
}
if (!msix_enabled(&proxy->pci_dev) &&
vdev->use_guest_notifier_mask &&
vdc->guest_notifier_mask) {
vdc->guest_notifier_mask(vdev, n, !assign);
}
return 0;
}
static bool virtio_pci_query_guest_notifiers(DeviceState *d)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
return msix_enabled(&proxy->pci_dev);
}
static int virtio_pci_set_guest_notifiers(DeviceState *d, int nvqs, bool assign)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
int r, n;
bool with_irqfd = msix_enabled(&proxy->pci_dev) &&
kvm_msi_via_irqfd_enabled();
nvqs = MIN(nvqs, VIRTIO_QUEUE_MAX);
/* When deassigning, pass a consistent nvqs value
* to avoid leaking notifiers.
*/
assert(assign || nvqs == proxy->nvqs_with_notifiers);
proxy->nvqs_with_notifiers = nvqs;
/* Must unset vector notifier while guest notifier is still assigned */
if ((proxy->vector_irqfd || k->guest_notifier_mask) && !assign) {
msix_unset_vector_notifiers(&proxy->pci_dev);
if (proxy->vector_irqfd) {
kvm_virtio_pci_vector_release(proxy, nvqs);
g_free(proxy->vector_irqfd);
proxy->vector_irqfd = NULL;
}
}
for (n = 0; n < nvqs; n++) {
if (!virtio_queue_get_num(vdev, n)) {
break;
}
r = virtio_pci_set_guest_notifier(d, n, assign, with_irqfd);
if (r < 0) {
goto assign_error;
}
}
/* Must set vector notifier after guest notifier has been assigned */
if ((with_irqfd || k->guest_notifier_mask) && assign) {
if (with_irqfd) {
proxy->vector_irqfd =
g_malloc0(sizeof(*proxy->vector_irqfd) *
msix_nr_vectors_allocated(&proxy->pci_dev));
r = kvm_virtio_pci_vector_use(proxy, nvqs);
if (r < 0) {
goto assign_error;
}
}
r = msix_set_vector_notifiers(&proxy->pci_dev,
virtio_pci_vector_unmask,
virtio_pci_vector_mask,
virtio_pci_vector_poll);
if (r < 0) {
goto notifiers_error;
}
}
return 0;
notifiers_error:
if (with_irqfd) {
assert(assign);
kvm_virtio_pci_vector_release(proxy, nvqs);
}
assign_error:
/* We get here on assignment failure. Recover by undoing for VQs 0 .. n. */
assert(assign);
while (--n >= 0) {
virtio_pci_set_guest_notifier(d, n, !assign, with_irqfd);
}
return r;
}
static int virtio_pci_set_host_notifier_mr(DeviceState *d, int n,
MemoryRegion *mr, bool assign)
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
int offset;
if (n >= VIRTIO_QUEUE_MAX || !virtio_pci_modern(proxy) ||
virtio_pci_queue_mem_mult(proxy) != memory_region_size(mr)) {
return -1;
}
if (assign) {
offset = virtio_pci_queue_mem_mult(proxy) * n;
memory_region_add_subregion_overlap(&proxy->notify.mr, offset, mr, 1);
} else {
memory_region_del_subregion(&proxy->notify.mr, mr);
}
return 0;
}
static void virtio_pci_vmstate_change(DeviceState *d, bool running)
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
{
VirtIOPCIProxy *proxy = to_virtio_pci_proxy(d);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
if (running) {
/* Old QEMU versions did not set bus master enable on status write.
* Detect DRIVER set and enable it.
*/
if ((proxy->flags & VIRTIO_PCI_FLAG_BUS_MASTER_BUG_MIGRATION) &&
(vdev->status & VIRTIO_CONFIG_S_DRIVER) &&
!(proxy->pci_dev.config[PCI_COMMAND] & PCI_COMMAND_MASTER)) {
pci_default_write_config(&proxy->pci_dev, PCI_COMMAND,
proxy->pci_dev.config[PCI_COMMAND] |
PCI_COMMAND_MASTER, 1);
}
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
virtio_pci_start_ioeventfd(proxy);
} else {
virtio-pci: Use ioeventfd for virtqueue notify Virtqueue notify is currently handled synchronously in userspace virtio. This prevents the vcpu from executing guest code while hardware emulation code handles the notify. On systems that support KVM, the ioeventfd mechanism can be used to make virtqueue notify a lightweight exit by deferring hardware emulation to the iothread and allowing the VM to continue execution. This model is similar to how vhost receives virtqueue notifies. The result of this change is improved performance for userspace virtio devices. Virtio-blk throughput increases especially for multithreaded scenarios and virtio-net transmit throughput increases substantially. Some virtio devices are known to have guest drivers which expect a notify to be processed synchronously and spin waiting for completion. For virtio-net, this also seems to interact with the guest stack in strange ways so that TCP throughput for small message sizes (~200bytes) is harmed. Only enable ioeventfd for virtio-blk for now. Care must be taken not to interfere with vhost-net, which uses host notifiers. If the set_host_notifier() API is used by a device virtio-pci will disable virtio-ioeventfd and let the device deal with host notifiers as it wishes. Finally, there used to be a limit of 6 KVM io bus devices inside the kernel. On such a kernel, don't use ioeventfd for virtqueue host notification since the limit is reached too easily. This ensures that existing vhost-net setups (which always use ioeventfd) have ioeventfds available so they can continue to work. After migration and on VM change state (running/paused) virtio-ioeventfd will enable/disable itself. * VIRTIO_CONFIG_S_DRIVER_OK -> enable virtio-ioeventfd * !VIRTIO_CONFIG_S_DRIVER_OK -> disable virtio-ioeventfd * virtio_pci_set_host_notifier() -> disable virtio-ioeventfd * vm_change_state(running=0) -> disable virtio-ioeventfd * vm_change_state(running=1) -> enable virtio-ioeventfd Signed-off-by: Stefan Hajnoczi <stefanha@linux.vnet.ibm.com> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2010-12-17 15:01:50 +03:00
virtio_pci_stop_ioeventfd(proxy);
}
}
/*
* virtio-pci: This is the PCIDevice which has a virtio-pci-bus.
*/
static int virtio_pci_query_nvectors(DeviceState *d)
{
VirtIOPCIProxy *proxy = VIRTIO_PCI(d);
return proxy->nvectors;
}
static AddressSpace *virtio_pci_get_dma_as(DeviceState *d)
{
VirtIOPCIProxy *proxy = VIRTIO_PCI(d);
PCIDevice *dev = &proxy->pci_dev;
return pci_get_address_space(dev);
}
static int virtio_pci_add_mem_cap(VirtIOPCIProxy *proxy,
struct virtio_pci_cap *cap)
{
PCIDevice *dev = &proxy->pci_dev;
int offset;
offset = pci_add_capability(dev, PCI_CAP_ID_VNDR, 0,
cap->cap_len, &error_abort);
assert(cap->cap_len >= sizeof *cap);
memcpy(dev->config + offset + PCI_CAP_FLAGS, &cap->cap_len,
cap->cap_len - PCI_CAP_FLAGS);
return offset;
}
static uint64_t virtio_pci_common_read(void *opaque, hwaddr addr,
unsigned size)
{
VirtIOPCIProxy *proxy = opaque;
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
uint32_t val = 0;
int i;
switch (addr) {
case VIRTIO_PCI_COMMON_DFSELECT:
val = proxy->dfselect;
break;
case VIRTIO_PCI_COMMON_DF:
if (proxy->dfselect <= 1) {
VirtioDeviceClass *vdc = VIRTIO_DEVICE_GET_CLASS(vdev);
val = (vdev->host_features & ~vdc->legacy_features) >>
(32 * proxy->dfselect);
}
break;
case VIRTIO_PCI_COMMON_GFSELECT:
val = proxy->gfselect;
break;
case VIRTIO_PCI_COMMON_GF:
if (proxy->gfselect < ARRAY_SIZE(proxy->guest_features)) {
val = proxy->guest_features[proxy->gfselect];
}
break;
case VIRTIO_PCI_COMMON_MSIX:
val = vdev->config_vector;
break;
case VIRTIO_PCI_COMMON_NUMQ:
for (i = 0; i < VIRTIO_QUEUE_MAX; ++i) {
if (virtio_queue_get_num(vdev, i)) {
val = i + 1;
}
}
break;
case VIRTIO_PCI_COMMON_STATUS:
val = vdev->status;
break;
case VIRTIO_PCI_COMMON_CFGGENERATION:
val = vdev->generation;
break;
case VIRTIO_PCI_COMMON_Q_SELECT:
val = vdev->queue_sel;
break;
case VIRTIO_PCI_COMMON_Q_SIZE:
val = virtio_queue_get_num(vdev, vdev->queue_sel);
break;
case VIRTIO_PCI_COMMON_Q_MSIX:
val = virtio_queue_vector(vdev, vdev->queue_sel);
break;
case VIRTIO_PCI_COMMON_Q_ENABLE:
val = proxy->vqs[vdev->queue_sel].enabled;
break;
case VIRTIO_PCI_COMMON_Q_NOFF:
/* Simply map queues in order */
val = vdev->queue_sel;
break;
case VIRTIO_PCI_COMMON_Q_DESCLO:
val = proxy->vqs[vdev->queue_sel].desc[0];
break;
case VIRTIO_PCI_COMMON_Q_DESCHI:
val = proxy->vqs[vdev->queue_sel].desc[1];
break;
case VIRTIO_PCI_COMMON_Q_AVAILLO:
val = proxy->vqs[vdev->queue_sel].avail[0];
break;
case VIRTIO_PCI_COMMON_Q_AVAILHI:
val = proxy->vqs[vdev->queue_sel].avail[1];
break;
case VIRTIO_PCI_COMMON_Q_USEDLO:
val = proxy->vqs[vdev->queue_sel].used[0];
break;
case VIRTIO_PCI_COMMON_Q_USEDHI:
val = proxy->vqs[vdev->queue_sel].used[1];
break;
default:
val = 0;
}
return val;
}
static void virtio_pci_common_write(void *opaque, hwaddr addr,
uint64_t val, unsigned size)
{
VirtIOPCIProxy *proxy = opaque;
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
switch (addr) {
case VIRTIO_PCI_COMMON_DFSELECT:
proxy->dfselect = val;
break;
case VIRTIO_PCI_COMMON_GFSELECT:
proxy->gfselect = val;
break;
case VIRTIO_PCI_COMMON_GF:
if (proxy->gfselect < ARRAY_SIZE(proxy->guest_features)) {
proxy->guest_features[proxy->gfselect] = val;
virtio_set_features(vdev,
(((uint64_t)proxy->guest_features[1]) << 32) |
proxy->guest_features[0]);
}
break;
case VIRTIO_PCI_COMMON_MSIX:
msix_vector_unuse(&proxy->pci_dev, vdev->config_vector);
/* Make it possible for guest to discover an error took place. */
if (msix_vector_use(&proxy->pci_dev, val) < 0) {
val = VIRTIO_NO_VECTOR;
}
vdev->config_vector = val;
break;
case VIRTIO_PCI_COMMON_STATUS:
if (!(val & VIRTIO_CONFIG_S_DRIVER_OK)) {
virtio_pci_stop_ioeventfd(proxy);
}
virtio_set_status(vdev, val & 0xFF);
if (val & VIRTIO_CONFIG_S_DRIVER_OK) {
virtio_pci_start_ioeventfd(proxy);
}
if (vdev->status == 0) {
virtio_pci_reset(DEVICE(proxy));
}
break;
case VIRTIO_PCI_COMMON_Q_SELECT:
if (val < VIRTIO_QUEUE_MAX) {
vdev->queue_sel = val;
}
break;
case VIRTIO_PCI_COMMON_Q_SIZE:
proxy->vqs[vdev->queue_sel].num = val;
break;
case VIRTIO_PCI_COMMON_Q_MSIX:
msix_vector_unuse(&proxy->pci_dev,
virtio_queue_vector(vdev, vdev->queue_sel));
/* Make it possible for guest to discover an error took place. */
if (msix_vector_use(&proxy->pci_dev, val) < 0) {
val = VIRTIO_NO_VECTOR;
}
virtio_queue_set_vector(vdev, vdev->queue_sel, val);
break;
case VIRTIO_PCI_COMMON_Q_ENABLE:
virtio_queue_set_num(vdev, vdev->queue_sel,
proxy->vqs[vdev->queue_sel].num);
virtio_queue_set_rings(vdev, vdev->queue_sel,
((uint64_t)proxy->vqs[vdev->queue_sel].desc[1]) << 32 |
proxy->vqs[vdev->queue_sel].desc[0],
((uint64_t)proxy->vqs[vdev->queue_sel].avail[1]) << 32 |
proxy->vqs[vdev->queue_sel].avail[0],
((uint64_t)proxy->vqs[vdev->queue_sel].used[1]) << 32 |
proxy->vqs[vdev->queue_sel].used[0]);
proxy->vqs[vdev->queue_sel].enabled = 1;
break;
case VIRTIO_PCI_COMMON_Q_DESCLO:
proxy->vqs[vdev->queue_sel].desc[0] = val;
break;
case VIRTIO_PCI_COMMON_Q_DESCHI:
proxy->vqs[vdev->queue_sel].desc[1] = val;
break;
case VIRTIO_PCI_COMMON_Q_AVAILLO:
proxy->vqs[vdev->queue_sel].avail[0] = val;
break;
case VIRTIO_PCI_COMMON_Q_AVAILHI:
proxy->vqs[vdev->queue_sel].avail[1] = val;
break;
case VIRTIO_PCI_COMMON_Q_USEDLO:
proxy->vqs[vdev->queue_sel].used[0] = val;
break;
case VIRTIO_PCI_COMMON_Q_USEDHI:
proxy->vqs[vdev->queue_sel].used[1] = val;
break;
default:
break;
}
}
static uint64_t virtio_pci_notify_read(void *opaque, hwaddr addr,
unsigned size)
{
return 0;
}
static void virtio_pci_notify_write(void *opaque, hwaddr addr,
uint64_t val, unsigned size)
{
VirtIODevice *vdev = opaque;
VirtIOPCIProxy *proxy = VIRTIO_PCI(DEVICE(vdev)->parent_bus->parent);
unsigned queue = addr / virtio_pci_queue_mem_mult(proxy);
if (queue < VIRTIO_QUEUE_MAX) {
virtio_queue_notify(vdev, queue);
}
}
static void virtio_pci_notify_write_pio(void *opaque, hwaddr addr,
uint64_t val, unsigned size)
{
VirtIODevice *vdev = opaque;
unsigned queue = val;
if (queue < VIRTIO_QUEUE_MAX) {
virtio_queue_notify(vdev, queue);
}
}
static uint64_t virtio_pci_isr_read(void *opaque, hwaddr addr,
unsigned size)
{
VirtIOPCIProxy *proxy = opaque;
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
uint64_t val = atomic_xchg(&vdev->isr, 0);
pci_irq_deassert(&proxy->pci_dev);
return val;
}
static void virtio_pci_isr_write(void *opaque, hwaddr addr,
uint64_t val, unsigned size)
{
}
static uint64_t virtio_pci_device_read(void *opaque, hwaddr addr,
unsigned size)
{
VirtIODevice *vdev = opaque;
uint64_t val = 0;
switch (size) {
case 1:
val = virtio_config_modern_readb(vdev, addr);
break;
case 2:
val = virtio_config_modern_readw(vdev, addr);
break;
case 4:
val = virtio_config_modern_readl(vdev, addr);
break;
}
return val;
}
static void virtio_pci_device_write(void *opaque, hwaddr addr,
uint64_t val, unsigned size)
{
VirtIODevice *vdev = opaque;
switch (size) {
case 1:
virtio_config_modern_writeb(vdev, addr, val);
break;
case 2:
virtio_config_modern_writew(vdev, addr, val);
break;
case 4:
virtio_config_modern_writel(vdev, addr, val);
break;
}
}
static void virtio_pci_modern_regions_init(VirtIOPCIProxy *proxy)
{
static const MemoryRegionOps common_ops = {
.read = virtio_pci_common_read,
.write = virtio_pci_common_write,
.impl = {
.min_access_size = 1,
.max_access_size = 4,
},
.endianness = DEVICE_LITTLE_ENDIAN,
};
static const MemoryRegionOps isr_ops = {
.read = virtio_pci_isr_read,
.write = virtio_pci_isr_write,
.impl = {
.min_access_size = 1,
.max_access_size = 4,
},
.endianness = DEVICE_LITTLE_ENDIAN,
};
static const MemoryRegionOps device_ops = {
.read = virtio_pci_device_read,
.write = virtio_pci_device_write,
.impl = {
.min_access_size = 1,
.max_access_size = 4,
},
.endianness = DEVICE_LITTLE_ENDIAN,
};
static const MemoryRegionOps notify_ops = {
.read = virtio_pci_notify_read,
.write = virtio_pci_notify_write,
.impl = {
.min_access_size = 1,
.max_access_size = 4,
},
.endianness = DEVICE_LITTLE_ENDIAN,
};
static const MemoryRegionOps notify_pio_ops = {
.read = virtio_pci_notify_read,
.write = virtio_pci_notify_write_pio,
.impl = {
.min_access_size = 1,
.max_access_size = 4,
},
.endianness = DEVICE_LITTLE_ENDIAN,
};
memory_region_init_io(&proxy->common.mr, OBJECT(proxy),
&common_ops,
proxy,
"virtio-pci-common",
proxy->common.size);
memory_region_init_io(&proxy->isr.mr, OBJECT(proxy),
&isr_ops,
proxy,
"virtio-pci-isr",
proxy->isr.size);
memory_region_init_io(&proxy->device.mr, OBJECT(proxy),
&device_ops,
virtio_bus_get_device(&proxy->bus),
"virtio-pci-device",
proxy->device.size);
memory_region_init_io(&proxy->notify.mr, OBJECT(proxy),
&notify_ops,
virtio_bus_get_device(&proxy->bus),
"virtio-pci-notify",
proxy->notify.size);
memory_region_init_io(&proxy->notify_pio.mr, OBJECT(proxy),
&notify_pio_ops,
virtio_bus_get_device(&proxy->bus),
"virtio-pci-notify-pio",
proxy->notify_pio.size);
}
static void virtio_pci_modern_region_map(VirtIOPCIProxy *proxy,
VirtIOPCIRegion *region,
struct virtio_pci_cap *cap,
MemoryRegion *mr,
uint8_t bar)
{
memory_region_add_subregion(mr, region->offset, &region->mr);
cap->cfg_type = region->type;
cap->bar = bar;
cap->offset = cpu_to_le32(region->offset);
cap->length = cpu_to_le32(region->size);
virtio_pci_add_mem_cap(proxy, cap);
}
static void virtio_pci_modern_mem_region_map(VirtIOPCIProxy *proxy,
VirtIOPCIRegion *region,
struct virtio_pci_cap *cap)
{
virtio_pci_modern_region_map(proxy, region, cap,
&proxy->modern_bar, proxy->modern_mem_bar_idx);
}
static void virtio_pci_modern_io_region_map(VirtIOPCIProxy *proxy,
VirtIOPCIRegion *region,
struct virtio_pci_cap *cap)
{
virtio_pci_modern_region_map(proxy, region, cap,
&proxy->io_bar, proxy->modern_io_bar_idx);
}
static void virtio_pci_modern_mem_region_unmap(VirtIOPCIProxy *proxy,
VirtIOPCIRegion *region)
{
memory_region_del_subregion(&proxy->modern_bar,
&region->mr);
}
static void virtio_pci_modern_io_region_unmap(VirtIOPCIProxy *proxy,
VirtIOPCIRegion *region)
{
memory_region_del_subregion(&proxy->io_bar,
&region->mr);
}
static void virtio_pci_pre_plugged(DeviceState *d, Error **errp)
{
VirtIOPCIProxy *proxy = VIRTIO_PCI(d);
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
if (virtio_pci_modern(proxy)) {
virtio_add_feature(&vdev->host_features, VIRTIO_F_VERSION_1);
}
virtio_add_feature(&vdev->host_features, VIRTIO_F_BAD_FEATURE);
}
/* This is called by virtio-bus just after the device is plugged. */
static void virtio_pci_device_plugged(DeviceState *d, Error **errp)
{
VirtIOPCIProxy *proxy = VIRTIO_PCI(d);
VirtioBusState *bus = &proxy->bus;
bool legacy = virtio_pci_legacy(proxy);
bool modern;
bool modern_pio = proxy->flags & VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY;
uint8_t *config;
uint32_t size;
VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
/*
* Virtio capabilities present without
* VIRTIO_F_VERSION_1 confuses guests
*/
virtio-pci: Fix cross-version migration with older machines This patch fixes a cross-version migration regression introduced by commit d1b4259f ("virtio-bus: Plug devices after features are negotiated"). The problem is encountered when host's vhost backend does not support VIRTIO_F_VERSION_1, and migration is initiated from a v2.7 or prior machine with virtio-pci modern capabilities enabled to a v2.8 machine. In this case, modern capabilities get exposed to the guest by the source, whereas the target will detect version 1 is not supported so will only expose legacy capabilities. The problem is fixed by introducing a new "x-ignore-backend-features" property, which is set in v2.7 and prior compatibility modes. Doing this, v2.7 machine keeps its broken behaviour (enabling modern while version is not supported), and newer machines will behave correctly. Reported-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Marcel Apfelbaum <marcel@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Tested-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com> Message-id: 20161214163035.3297-1-maxime.coquelin@redhat.com Suggested-by: Stefan Hajnoczi <stefanha@redhat.com> Cc: Michael S. Tsirkin <mst@redhat.com> Cc: Cornelia Huck <cornelia.huck@de.ibm.com> Cc: Dr. David Alan Gilbert <dgilbert@redhat.com> Reviewed-by: Marcel Apfelbaum <marcel@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Tested-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2016-12-14 19:30:35 +03:00
if (!proxy->ignore_backend_features &&
!virtio_has_feature(vdev->host_features, VIRTIO_F_VERSION_1)) {
virtio_pci_disable_modern(proxy);
if (!legacy) {
error_setg(errp, "Device doesn't support modern mode, and legacy"
" mode is disabled");
error_append_hint(errp, "Set disable-legacy to off\n");
return;
}
}
modern = virtio_pci_modern(proxy);
config = proxy->pci_dev.config;
if (proxy->class_code) {
pci_config_set_class(config, proxy->class_code);
}
if (legacy) {
if (virtio_host_has_feature(vdev, VIRTIO_F_IOMMU_PLATFORM)) {
error_setg(errp, "VIRTIO_F_IOMMU_PLATFORM was supported by"
" neither legacy nor transitional device");
return ;
}
/*
* Legacy and transitional devices use specific subsystem IDs.
* Note that the subsystem vendor ID (config + PCI_SUBSYSTEM_VENDOR_ID)
* is set to PCI_SUBVENDOR_ID_REDHAT_QUMRANET by default.
*/
pci_set_word(config + PCI_SUBSYSTEM_ID, virtio_bus_get_vdev_id(bus));
} else {
/* pure virtio-1.0 */
pci_set_word(config + PCI_VENDOR_ID,
PCI_VENDOR_ID_REDHAT_QUMRANET);
pci_set_word(config + PCI_DEVICE_ID,
0x1040 + virtio_bus_get_vdev_id(bus));
pci_config_set_revision(config, 1);
}
config[PCI_INTERRUPT_PIN] = 1;
if (modern) {
struct virtio_pci_cap cap = {
.cap_len = sizeof cap,
};
struct virtio_pci_notify_cap notify = {
.cap.cap_len = sizeof notify,
.notify_off_multiplier =
cpu_to_le32(virtio_pci_queue_mem_mult(proxy)),
};
struct virtio_pci_cfg_cap cfg = {
.cap.cap_len = sizeof cfg,
.cap.cfg_type = VIRTIO_PCI_CAP_PCI_CFG,
};
struct virtio_pci_notify_cap notify_pio = {
.cap.cap_len = sizeof notify,
.notify_off_multiplier = cpu_to_le32(0x0),
};
struct virtio_pci_cfg_cap *cfg_mask;
virtio_pci_modern_regions_init(proxy);
virtio_pci_modern_mem_region_map(proxy, &proxy->common, &cap);
virtio_pci_modern_mem_region_map(proxy, &proxy->isr, &cap);
virtio_pci_modern_mem_region_map(proxy, &proxy->device, &cap);
virtio_pci_modern_mem_region_map(proxy, &proxy->notify, &notify.cap);
if (modern_pio) {
memory_region_init(&proxy->io_bar, OBJECT(proxy),
"virtio-pci-io", 0x4);
pci_register_bar(&proxy->pci_dev, proxy->modern_io_bar_idx,
PCI_BASE_ADDRESS_SPACE_IO, &proxy->io_bar);
virtio_pci_modern_io_region_map(proxy, &proxy->notify_pio,
&notify_pio.cap);
}
pci_register_bar(&proxy->pci_dev, proxy->modern_mem_bar_idx,
PCI_BASE_ADDRESS_SPACE_MEMORY |
PCI_BASE_ADDRESS_MEM_PREFETCH |
PCI_BASE_ADDRESS_MEM_TYPE_64,
&proxy->modern_bar);
proxy->config_cap = virtio_pci_add_mem_cap(proxy, &cfg.cap);
cfg_mask = (void *)(proxy->pci_dev.wmask + proxy->config_cap);
pci_set_byte(&cfg_mask->cap.bar, ~0x0);
pci_set_long((uint8_t *)&cfg_mask->cap.offset, ~0x0);
pci_set_long((uint8_t *)&cfg_mask->cap.length, ~0x0);
pci_set_long(cfg_mask->pci_cfg_data, ~0x0);
}
if (proxy->nvectors) {
int err = msix_init_exclusive_bar(&proxy->pci_dev, proxy->nvectors,
proxy->msix_bar_idx, NULL);
if (err) {
/* Notice when a system that supports MSIx can't initialize it */
if (err != -ENOTSUP) {
warn_report("unable to init msix vectors to %" PRIu32,
proxy->nvectors);
}
proxy->nvectors = 0;
}
}
proxy->pci_dev.config_write = virtio_write_config;
proxy->pci_dev.config_read = virtio_read_config;
if (legacy) {
size = VIRTIO_PCI_REGION_SIZE(&proxy->pci_dev)
+ virtio_bus_get_vdev_config_len(bus);
size = pow2ceil(size);
memory_region_init_io(&proxy->bar, OBJECT(proxy),
&virtio_pci_config_ops,
proxy, "virtio-pci", size);
pci_register_bar(&proxy->pci_dev, proxy->legacy_io_bar_idx,
PCI_BASE_ADDRESS_SPACE_IO, &proxy->bar);
}
}
static void virtio_pci_device_unplugged(DeviceState *d)
{
VirtIOPCIProxy *proxy = VIRTIO_PCI(d);
bool modern = virtio_pci_modern(proxy);
bool modern_pio = proxy->flags & VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY;
virtio_pci_stop_ioeventfd(proxy);
if (modern) {
virtio_pci_modern_mem_region_unmap(proxy, &proxy->common);
virtio_pci_modern_mem_region_unmap(proxy, &proxy->isr);
virtio_pci_modern_mem_region_unmap(proxy, &proxy->device);
virtio_pci_modern_mem_region_unmap(proxy, &proxy->notify);
if (modern_pio) {
virtio_pci_modern_io_region_unmap(proxy, &proxy->notify_pio);
}
}
}
static void virtio_pci_realize(PCIDevice *pci_dev, Error **errp)
{
VirtIOPCIProxy *proxy = VIRTIO_PCI(pci_dev);
VirtioPCIClass *k = VIRTIO_PCI_GET_CLASS(pci_dev);
bool pcie_port = pci_bus_is_express(pci_get_bus(pci_dev)) &&
!pci_bus_is_root(pci_get_bus(pci_dev));
if (kvm_enabled() && !kvm_has_many_ioeventfds()) {
proxy->flags &= ~VIRTIO_PCI_FLAG_USE_IOEVENTFD;
}
/*
* virtio pci bar layout used by default.
* subclasses can re-arrange things if needed.
*
* region 0 -- virtio legacy io bar
* region 1 -- msi-x bar
* region 4+5 -- virtio modern memory (64bit) bar
*
*/
proxy->legacy_io_bar_idx = 0;
proxy->msix_bar_idx = 1;
proxy->modern_io_bar_idx = 2;
proxy->modern_mem_bar_idx = 4;
proxy->common.offset = 0x0;
proxy->common.size = 0x1000;
proxy->common.type = VIRTIO_PCI_CAP_COMMON_CFG;
proxy->isr.offset = 0x1000;
proxy->isr.size = 0x1000;
proxy->isr.type = VIRTIO_PCI_CAP_ISR_CFG;
proxy->device.offset = 0x2000;
proxy->device.size = 0x1000;
proxy->device.type = VIRTIO_PCI_CAP_DEVICE_CFG;
proxy->notify.offset = 0x3000;
proxy->notify.size = virtio_pci_queue_mem_mult(proxy) * VIRTIO_QUEUE_MAX;
proxy->notify.type = VIRTIO_PCI_CAP_NOTIFY_CFG;
proxy->notify_pio.offset = 0x0;
proxy->notify_pio.size = 0x4;
proxy->notify_pio.type = VIRTIO_PCI_CAP_NOTIFY_CFG;
/* subclasses can enforce modern, so do this unconditionally */
memory_region_init(&proxy->modern_bar, OBJECT(proxy), "virtio-pci",
/* PCI BAR regions must be powers of 2 */
pow2ceil(proxy->notify.offset + proxy->notify.size));
if ((proxy->disable_legacy == ON_OFF_AUTO_ON) ||
((proxy->disable_legacy == ON_OFF_AUTO_AUTO) && pcie_port)) {
if (proxy->disable_modern) {
error_setg(errp, "device cannot work as neither modern nor "
"legacy mode is enabled");
error_append_hint(errp, "Set either disable-modern or "
"disable-legacy to off\n");
return;
}
proxy->mode = VIRTIO_PCI_MODE_MODERN;
} else {
if (proxy->disable_modern) {
proxy->mode = VIRTIO_PCI_MODE_LEGACY;
} else {
proxy->mode = VIRTIO_PCI_MODE_TRANSITIONAL;
}
}
if (pcie_port && pci_is_express(pci_dev)) {
int pos;
pos = pcie_endpoint_cap_init(pci_dev, 0);
assert(pos > 0);
pos = pci_add_capability(pci_dev, PCI_CAP_ID_PM, 0,
PCI_PM_SIZEOF, errp);
if (pos < 0) {
return;
}
pci_dev->exp.pm_cap = pos;
/*
* Indicates that this function complies with revision 1.2 of the
* PCI Power Management Interface Specification.
*/
pci_set_word(pci_dev->config + pos + PCI_PM_PMC, 0x3);
if (proxy->flags & VIRTIO_PCI_FLAG_INIT_DEVERR) {
/* Init error enabling flags */
pcie_cap_deverr_init(pci_dev);
}
if (proxy->flags & VIRTIO_PCI_FLAG_INIT_LNKCTL) {
/* Init Link Control Register */
pcie_cap_lnkctl_init(pci_dev);
}
if (proxy->flags & VIRTIO_PCI_FLAG_INIT_PM) {
/* Init Power Management Control Register */
pci_set_word(pci_dev->wmask + pos + PCI_PM_CTRL,
PCI_PM_CTRL_STATE_MASK);
}
if (proxy->flags & VIRTIO_PCI_FLAG_ATS) {
pcie_ats_init(pci_dev, 256);
}
} else {
/*
* make future invocations of pci_is_express() return false
* and pci_config_size() return PCI_CONFIG_SPACE_SIZE.
*/
pci_dev->cap_present &= ~QEMU_PCI_CAP_EXPRESS;
}
virtio_pci_bus_new(&proxy->bus, sizeof(proxy->bus), proxy);
if (k->realize) {
k->realize(proxy, errp);
}
}
static void virtio_pci_exit(PCIDevice *pci_dev)
{
msix_uninit_exclusive_bar(pci_dev);
}
static void virtio_pci_reset(DeviceState *qdev)
{
VirtIOPCIProxy *proxy = VIRTIO_PCI(qdev);
VirtioBusState *bus = VIRTIO_BUS(&proxy->bus);
PCIDevice *dev = PCI_DEVICE(qdev);
int i;
virtio_pci_stop_ioeventfd(proxy);
virtio_bus_reset(bus);
msix_unuse_all_vectors(&proxy->pci_dev);
for (i = 0; i < VIRTIO_QUEUE_MAX; i++) {
proxy->vqs[i].enabled = 0;
proxy->vqs[i].num = 0;
proxy->vqs[i].desc[0] = proxy->vqs[i].desc[1] = 0;
proxy->vqs[i].avail[0] = proxy->vqs[i].avail[1] = 0;
proxy->vqs[i].used[0] = proxy->vqs[i].used[1] = 0;
}
if (pci_is_express(dev)) {
pcie_cap_deverr_reset(dev);
pcie_cap_lnkctl_reset(dev);
pci_set_word(dev->config + dev->exp.pm_cap + PCI_PM_CTRL, 0);
}
}
static Property virtio_pci_properties[] = {
DEFINE_PROP_BIT("virtio-pci-bus-master-bug-migration", VirtIOPCIProxy, flags,
VIRTIO_PCI_FLAG_BUS_MASTER_BUG_MIGRATION_BIT, false),
DEFINE_PROP_BIT("migrate-extra", VirtIOPCIProxy, flags,
VIRTIO_PCI_FLAG_MIGRATE_EXTRA_BIT, true),
DEFINE_PROP_BIT("modern-pio-notify", VirtIOPCIProxy, flags,
VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY_BIT, false),
DEFINE_PROP_BIT("x-disable-pcie", VirtIOPCIProxy, flags,
VIRTIO_PCI_FLAG_DISABLE_PCIE_BIT, false),
DEFINE_PROP_BIT("page-per-vq", VirtIOPCIProxy, flags,
VIRTIO_PCI_FLAG_PAGE_PER_VQ_BIT, false),
virtio-pci: Fix cross-version migration with older machines This patch fixes a cross-version migration regression introduced by commit d1b4259f ("virtio-bus: Plug devices after features are negotiated"). The problem is encountered when host's vhost backend does not support VIRTIO_F_VERSION_1, and migration is initiated from a v2.7 or prior machine with virtio-pci modern capabilities enabled to a v2.8 machine. In this case, modern capabilities get exposed to the guest by the source, whereas the target will detect version 1 is not supported so will only expose legacy capabilities. The problem is fixed by introducing a new "x-ignore-backend-features" property, which is set in v2.7 and prior compatibility modes. Doing this, v2.7 machine keeps its broken behaviour (enabling modern while version is not supported), and newer machines will behave correctly. Reported-by: Michael Roth <mdroth@linux.vnet.ibm.com> Reviewed-by: Marcel Apfelbaum <marcel@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Michael S. Tsirkin <mst@redhat.com> Tested-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com> Message-id: 20161214163035.3297-1-maxime.coquelin@redhat.com Suggested-by: Stefan Hajnoczi <stefanha@redhat.com> Cc: Michael S. Tsirkin <mst@redhat.com> Cc: Cornelia Huck <cornelia.huck@de.ibm.com> Cc: Dr. David Alan Gilbert <dgilbert@redhat.com> Reviewed-by: Marcel Apfelbaum <marcel@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Tested-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
2016-12-14 19:30:35 +03:00
DEFINE_PROP_BOOL("x-ignore-backend-features", VirtIOPCIProxy,
ignore_backend_features, false),
DEFINE_PROP_BIT("ats", VirtIOPCIProxy, flags,
VIRTIO_PCI_FLAG_ATS_BIT, false),
DEFINE_PROP_BIT("x-pcie-deverr-init", VirtIOPCIProxy, flags,
VIRTIO_PCI_FLAG_INIT_DEVERR_BIT, true),
DEFINE_PROP_BIT("x-pcie-lnkctl-init", VirtIOPCIProxy, flags,
VIRTIO_PCI_FLAG_INIT_LNKCTL_BIT, true),
DEFINE_PROP_BIT("x-pcie-pm-init", VirtIOPCIProxy, flags,
VIRTIO_PCI_FLAG_INIT_PM_BIT, true),
DEFINE_PROP_END_OF_LIST(),
};
static void virtio_pci_dc_realize(DeviceState *qdev, Error **errp)
{
VirtioPCIClass *vpciklass = VIRTIO_PCI_GET_CLASS(qdev);
VirtIOPCIProxy *proxy = VIRTIO_PCI(qdev);
PCIDevice *pci_dev = &proxy->pci_dev;
if (!(proxy->flags & VIRTIO_PCI_FLAG_DISABLE_PCIE) &&
virtio_pci_modern(proxy)) {
pci_dev->cap_present |= QEMU_PCI_CAP_EXPRESS;
}
vpciklass->parent_dc_realize(qdev, errp);
}
static void virtio_pci_class_init(ObjectClass *klass, void *data)
{
DeviceClass *dc = DEVICE_CLASS(klass);
PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
VirtioPCIClass *vpciklass = VIRTIO_PCI_CLASS(klass);
dc->props = virtio_pci_properties;
k->realize = virtio_pci_realize;
k->exit = virtio_pci_exit;
k->vendor_id = PCI_VENDOR_ID_REDHAT_QUMRANET;
k->revision = VIRTIO_PCI_ABI_VERSION;
k->class_id = PCI_CLASS_OTHERS;
device_class_set_parent_realize(dc, virtio_pci_dc_realize,
&vpciklass->parent_dc_realize);
dc->reset = virtio_pci_reset;
}
static const TypeInfo virtio_pci_info = {
.name = TYPE_VIRTIO_PCI,
.parent = TYPE_PCI_DEVICE,
.instance_size = sizeof(VirtIOPCIProxy),
.class_init = virtio_pci_class_init,
.class_size = sizeof(VirtioPCIClass),
.abstract = true,
};
static Property virtio_pci_generic_properties[] = {
DEFINE_PROP_ON_OFF_AUTO("disable-legacy", VirtIOPCIProxy, disable_legacy,
ON_OFF_AUTO_AUTO),
DEFINE_PROP_BOOL("disable-modern", VirtIOPCIProxy, disable_modern, false),
DEFINE_PROP_END_OF_LIST(),
};
static void virtio_pci_base_class_init(ObjectClass *klass, void *data)
{
const VirtioPCIDeviceTypeInfo *t = data;
if (t->class_init) {
t->class_init(klass, NULL);
}
}
static void virtio_pci_generic_class_init(ObjectClass *klass, void *data)
{
DeviceClass *dc = DEVICE_CLASS(klass);
dc->props = virtio_pci_generic_properties;
}
static void virtio_pci_transitional_instance_init(Object *obj)
{
VirtIOPCIProxy *proxy = VIRTIO_PCI(obj);
proxy->disable_legacy = ON_OFF_AUTO_OFF;
proxy->disable_modern = false;
}
static void virtio_pci_non_transitional_instance_init(Object *obj)
{
VirtIOPCIProxy *proxy = VIRTIO_PCI(obj);
proxy->disable_legacy = ON_OFF_AUTO_ON;
proxy->disable_modern = false;
}
void virtio_pci_types_register(const VirtioPCIDeviceTypeInfo *t)
{
char *base_name = NULL;
TypeInfo base_type_info = {
.name = t->base_name,
.parent = t->parent ? t->parent : TYPE_VIRTIO_PCI,
.instance_size = t->instance_size,
.instance_init = t->instance_init,
.class_size = t->class_size,
.abstract = true,
.interfaces = t->interfaces,
};
TypeInfo generic_type_info = {
.name = t->generic_name,
.parent = base_type_info.name,
.class_init = virtio_pci_generic_class_init,
.interfaces = (InterfaceInfo[]) {
{ INTERFACE_PCIE_DEVICE },
{ INTERFACE_CONVENTIONAL_PCI_DEVICE },
{ }
},
};
if (!base_type_info.name) {
/* No base type -> register a single generic device type */
/* use intermediate %s-base-type to add generic device props */
base_name = g_strdup_printf("%s-base-type", t->generic_name);
base_type_info.name = base_name;
base_type_info.class_init = virtio_pci_generic_class_init;
generic_type_info.parent = base_name;
generic_type_info.class_init = virtio_pci_base_class_init;
generic_type_info.class_data = (void *)t;
assert(!t->non_transitional_name);
assert(!t->transitional_name);
} else {
base_type_info.class_init = virtio_pci_base_class_init;
base_type_info.class_data = (void *)t;
}
type_register(&base_type_info);
if (generic_type_info.name) {
type_register(&generic_type_info);
}
if (t->non_transitional_name) {
const TypeInfo non_transitional_type_info = {
.name = t->non_transitional_name,
.parent = base_type_info.name,
.instance_init = virtio_pci_non_transitional_instance_init,
.interfaces = (InterfaceInfo[]) {
{ INTERFACE_PCIE_DEVICE },
{ INTERFACE_CONVENTIONAL_PCI_DEVICE },
{ }
},
};
type_register(&non_transitional_type_info);
}
if (t->transitional_name) {
const TypeInfo transitional_type_info = {
.name = t->transitional_name,
.parent = base_type_info.name,
.instance_init = virtio_pci_transitional_instance_init,
.interfaces = (InterfaceInfo[]) {
/*
* Transitional virtio devices work only as Conventional PCI
* devices because they require PIO ports.
*/
{ INTERFACE_CONVENTIONAL_PCI_DEVICE },
{ }
},
};
type_register(&transitional_type_info);
}
g_free(base_name);
}
/* virtio-pci-bus */
static void virtio_pci_bus_new(VirtioBusState *bus, size_t bus_size,
VirtIOPCIProxy *dev)
{
DeviceState *qdev = DEVICE(dev);
char virtio_bus_name[] = "virtio-bus";
qbus_create_inplace(bus, bus_size, TYPE_VIRTIO_PCI_BUS, qdev,
virtio_bus_name);
}
static void virtio_pci_bus_class_init(ObjectClass *klass, void *data)
{
BusClass *bus_class = BUS_CLASS(klass);
VirtioBusClass *k = VIRTIO_BUS_CLASS(klass);
bus_class->max_dev = 1;
k->notify = virtio_pci_notify;
k->save_config = virtio_pci_save_config;
k->load_config = virtio_pci_load_config;
k->save_queue = virtio_pci_save_queue;
k->load_queue = virtio_pci_load_queue;
k->save_extra_state = virtio_pci_save_extra_state;
k->load_extra_state = virtio_pci_load_extra_state;
k->has_extra_state = virtio_pci_has_extra_state;
k->query_guest_notifiers = virtio_pci_query_guest_notifiers;
k->set_guest_notifiers = virtio_pci_set_guest_notifiers;
k->set_host_notifier_mr = virtio_pci_set_host_notifier_mr;
k->vmstate_change = virtio_pci_vmstate_change;
k->pre_plugged = virtio_pci_pre_plugged;
k->device_plugged = virtio_pci_device_plugged;
k->device_unplugged = virtio_pci_device_unplugged;
k->query_nvectors = virtio_pci_query_nvectors;
k->ioeventfd_enabled = virtio_pci_ioeventfd_enabled;
k->ioeventfd_assign = virtio_pci_ioeventfd_assign;
k->get_dma_as = virtio_pci_get_dma_as;
}
static const TypeInfo virtio_pci_bus_info = {
.name = TYPE_VIRTIO_PCI_BUS,
.parent = TYPE_VIRTIO_BUS,
.instance_size = sizeof(VirtioPCIBusState),
.class_init = virtio_pci_bus_class_init,
};
static void virtio_pci_register_types(void)
{
/* Base types: */
type_register_static(&virtio_pci_bus_info);
type_register_static(&virtio_pci_info);
}
type_init(virtio_pci_register_types)