[Qemu-devel] [PATCH v3 0/5] virtio: fix some issues of "started" and "start_on_kick" flag

[Qemu-devel] [PATCH v3 0/5] virtio: fix some issues of "started" and "start_on_kick" flag

[elohimes]

From: Xie Yongji <xieyongji@baidu.com>

We introduced two flags "started" and "start_on_kick" to indicate virtio
device's state before. But there still are some problems with them. So
we try to fixup them in this patchset.

The patch 1 introduces a "use-started" property to avoid a migration
issue under Greg Kurz's suggestion [1].

The patch 2 set "start_on_kick" flag for legacy devices.

The patch 3 fixes a regression bug that old guest is not able to boot with
vhost-user-blk device.

The patch 4,5 fix some problems with "started" and "start_on_kick" flag.

[1] https://lists.gnu.org/archive/html/qemu-devel/2019-05/msg06247.html

v3:
- change the order of patches
- Also disable "use-started" property by hw_compat_4_0

v2:
- Recalculate "start_on_kick" flag after migration instead of migrating
  it
- Set "start_on_kick" flag for legacy devices
- Avoid setting "started" flag when "use_started" property is true
- Set "use_started" to false by hw_compat_4_0_1 instead of hw_compat_4_0

Xie Yongji (5):
  virtio: add "use-started" property
  virtio: Set "start_on_kick" for legacy devices
  virtio: Set "start_on_kick" on virtio_set_features()
  virtio: Make sure we get correct state of device on
    handle_aio_output()
  virtio: Don't change "started" flag on virtio_vmstate_change()

 hw/block/vhost-user-blk.c  |  4 +--
 hw/core/machine.c          |  8 ++++--
 hw/virtio/virtio.c         | 53 ++++++++++++++++++++++----------------
 include/hw/virtio/virtio.h | 23 ++++++++++++++++-
 4 files changed, 61 insertions(+), 27 deletions(-)

-- 
2.17.1

[Qemu-devel] [PATCH v3 1/5] virtio: add "use-started" property

[elohimes]

From: Xie Yongji <xieyongji@baidu.com>

In order to avoid migration issues, we introduce a "use-started"
property to the base virtio device to indicate whether use
"started" flag or not. This property will be true by default and
set to false when machine type <= 4.0.1.

Suggested-by: Greg Kurz <groug@kaod.org>
Signed-off-by: Xie Yongji <xieyongji@baidu.com>
---
 hw/block/vhost-user-blk.c  |  4 ++--
 hw/core/machine.c          |  8 ++++++--
 hw/virtio/virtio.c         | 18 +++++++-----------
 include/hw/virtio/virtio.h | 21 +++++++++++++++++++++
 4 files changed, 36 insertions(+), 15 deletions(-)

diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 9cb61336a6..85bc4017e7 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -191,7 +191,7 @@ static void vhost_user_blk_stop(VirtIODevice *vdev)
 static void vhost_user_blk_set_status(VirtIODevice *vdev, uint8_t status)
 {
     VHostUserBlk *s = VHOST_USER_BLK(vdev);
-    bool should_start = vdev->started;
+    bool should_start = virtio_device_started(vdev, status);
     int ret;
 
     if (!vdev->vm_running) {
@@ -317,7 +317,7 @@ static int vhost_user_blk_connect(DeviceState *dev)
     }
 
     /* restore vhost state */
-    if (vdev->started) {
+    if (virtio_device_started(vdev, vdev->status)) {
         ret = vhost_user_blk_start(vdev);
         if (ret < 0) {
             error_report("vhost-user-blk: vhost start failed: %s",
diff --git a/hw/core/machine.c b/hw/core/machine.c
index f1a0f45f9c..60d1e27d2c 100644
--- a/hw/core/machine.c
+++ b/hw/core/machine.c
@@ -24,10 +24,14 @@
 #include "hw/pci/pci.h"
 #include "hw/mem/nvdimm.h"
 
-GlobalProperty hw_compat_4_0_1[] = {};
+GlobalProperty hw_compat_4_0_1[] = {
+    { "virtio-device", "use-started", "false" },
+};
 const size_t hw_compat_4_0_1_len = G_N_ELEMENTS(hw_compat_4_0_1);
 
-GlobalProperty hw_compat_4_0[] = {};
+GlobalProperty hw_compat_4_0[] = {
+    { "virtio-device", "use-started", "false" },
+};
 const size_t hw_compat_4_0_len = G_N_ELEMENTS(hw_compat_4_0);
 
 GlobalProperty hw_compat_3_1[] = {
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 07f4a64b48..19062fbb96 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -1162,10 +1162,8 @@ int virtio_set_status(VirtIODevice *vdev, uint8_t val)
             }
         }
     }
-    vdev->started = val & VIRTIO_CONFIG_S_DRIVER_OK;
-    if (unlikely(vdev->start_on_kick && vdev->started)) {
-        vdev->start_on_kick = false;
-    }
+
+    virtio_set_started(vdev, val & VIRTIO_CONFIG_S_DRIVER_OK);
 
     if (k->set_status) {
         k->set_status(vdev, val);
@@ -1536,8 +1534,7 @@ static bool virtio_queue_notify_aio_vq(VirtQueue *vq)
         ret = vq->handle_aio_output(vdev, vq);
 
         if (unlikely(vdev->start_on_kick)) {
-            vdev->started = true;
-            vdev->start_on_kick = false;
+            virtio_set_started(vdev, true);
         }
     }
 
@@ -1557,8 +1554,7 @@ static void virtio_queue_notify_vq(VirtQueue *vq)
         vq->handle_output(vdev, vq);
 
         if (unlikely(vdev->start_on_kick)) {
-            vdev->started = true;
-            vdev->start_on_kick = false;
+            virtio_set_started(vdev, true);
         }
     }
 }
@@ -1579,8 +1575,7 @@ void virtio_queue_notify(VirtIODevice *vdev, int n)
     }
 
     if (unlikely(vdev->start_on_kick)) {
-        vdev->started = true;
-        vdev->start_on_kick = false;
+        virtio_set_started(vdev, true);
     }
 }
 
@@ -2291,7 +2286,7 @@ static void virtio_vmstate_change(void *opaque, int running, RunState state)
     VirtIODevice *vdev = opaque;
     BusState *qbus = qdev_get_parent_bus(DEVICE(vdev));
     VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
-    bool backend_run = running && vdev->started;
+    bool backend_run = running && virtio_device_started(vdev, vdev->status);
     vdev->vm_running = running;
 
     if (backend_run) {
@@ -2669,6 +2664,7 @@ static void virtio_device_instance_finalize(Object *obj)
 
 static Property virtio_properties[] = {
     DEFINE_VIRTIO_COMMON_FEATURES(VirtIODevice, host_features),
+    DEFINE_PROP_BOOL("use-started", VirtIODevice, use_started, true),
     DEFINE_PROP_END_OF_LIST(),
 };
 
diff --git a/include/hw/virtio/virtio.h b/include/hw/virtio/virtio.h
index 27c0efc3d0..15d5366939 100644
--- a/include/hw/virtio/virtio.h
+++ b/include/hw/virtio/virtio.h
@@ -105,6 +105,7 @@ struct VirtIODevice
     uint16_t device_id;
     bool vm_running;
     bool broken; /* device in invalid state, needs reset */
+    bool use_started;
     bool started;
     bool start_on_kick; /* virtio 1.0 transitional devices support that */
     VMChangeStateEntry *vmstate;
@@ -351,4 +352,24 @@ static inline bool virtio_is_big_endian(VirtIODevice *vdev)
     /* Devices conforming to VIRTIO 1.0 or later are always LE. */
     return false;
 }
+
+static inline bool virtio_device_started(VirtIODevice *vdev, uint8_t status)
+{
+    if (vdev->use_started) {
+        return vdev->started;
+    }
+
+    return status & VIRTIO_CONFIG_S_DRIVER_OK;
+}
+
+static inline void virtio_set_started(VirtIODevice *vdev, bool started)
+{
+    if (started) {
+        vdev->start_on_kick = false;
+    }
+
+    if (vdev->use_started) {
+        vdev->started = started;
+    }
+}
 #endif
-- 
2.17.1

Re: [Qemu-devel] [PATCH v3 1/5] virtio: add "use-started" property

[Greg Kurz]

On Fri, 14 Jun 2019 17:31:17 +0800
elohimes@gmail.com wrote:

> From: Xie Yongji <xieyongji@baidu.com>
> 
> In order to avoid migration issues, we introduce a "use-started"
> property to the base virtio device to indicate whether use
> "started" flag or not. This property will be true by default and
> set to false when machine type <= 4.0.1.
> 
> Suggested-by: Greg Kurz <groug@kaod.org>
> Signed-off-by: Xie Yongji <xieyongji@baidu.com>
> ---
>  hw/block/vhost-user-blk.c  |  4 ++--
>  hw/core/machine.c          |  8 ++++++--

This patch conflicts with latest upstream changes to hw_compat_4_0_1[].

It seems you need to rebase. Also, I'm still not sure how we're supposed
to handle hw_compat_4_0_1[] versus hw_compat_4_0[]... nobody commented
on:

https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg00637.html
https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg00641.html

Maybe worth to sort that out before re-posting.

>  hw/virtio/virtio.c         | 18 +++++++-----------
>  include/hw/virtio/virtio.h | 21 +++++++++++++++++++++
>  4 files changed, 36 insertions(+), 15 deletions(-)
> 
> diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
> index 9cb61336a6..85bc4017e7 100644
> --- a/hw/block/vhost-user-blk.c
> +++ b/hw/block/vhost-user-blk.c
> @@ -191,7 +191,7 @@ static void vhost_user_blk_stop(VirtIODevice *vdev)
>  static void vhost_user_blk_set_status(VirtIODevice *vdev, uint8_t status)
>  {
>      VHostUserBlk *s = VHOST_USER_BLK(vdev);
> -    bool should_start = vdev->started;
> +    bool should_start = virtio_device_started(vdev, status);
>      int ret;
>  
>      if (!vdev->vm_running) {
> @@ -317,7 +317,7 @@ static int vhost_user_blk_connect(DeviceState *dev)
>      }
>  
>      /* restore vhost state */
> -    if (vdev->started) {
> +    if (virtio_device_started(vdev, vdev->status)) {
>          ret = vhost_user_blk_start(vdev);
>          if (ret < 0) {
>              error_report("vhost-user-blk: vhost start failed: %s",
> diff --git a/hw/core/machine.c b/hw/core/machine.c
> index f1a0f45f9c..60d1e27d2c 100644
> --- a/hw/core/machine.c
> +++ b/hw/core/machine.c
> @@ -24,10 +24,14 @@
>  #include "hw/pci/pci.h"
>  #include "hw/mem/nvdimm.h"
>  
> -GlobalProperty hw_compat_4_0_1[] = {};
> +GlobalProperty hw_compat_4_0_1[] = {
> +    { "virtio-device", "use-started", "false" },
> +};
>  const size_t hw_compat_4_0_1_len = G_N_ELEMENTS(hw_compat_4_0_1);
>  
> -GlobalProperty hw_compat_4_0[] = {};
> +GlobalProperty hw_compat_4_0[] = {
> +    { "virtio-device", "use-started", "false" },
> +};
>  const size_t hw_compat_4_0_len = G_N_ELEMENTS(hw_compat_4_0);
>  
>  GlobalProperty hw_compat_3_1[] = {
> diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
> index 07f4a64b48..19062fbb96 100644
> --- a/hw/virtio/virtio.c
> +++ b/hw/virtio/virtio.c
> @@ -1162,10 +1162,8 @@ int virtio_set_status(VirtIODevice *vdev, uint8_t val)
>              }
>          }
>      }
> -    vdev->started = val & VIRTIO_CONFIG_S_DRIVER_OK;
> -    if (unlikely(vdev->start_on_kick && vdev->started)) {
> -        vdev->start_on_kick = false;
> -    }
> +
> +    virtio_set_started(vdev, val & VIRTIO_CONFIG_S_DRIVER_OK);
>  
>      if (k->set_status) {
>          k->set_status(vdev, val);
> @@ -1536,8 +1534,7 @@ static bool virtio_queue_notify_aio_vq(VirtQueue *vq)
>          ret = vq->handle_aio_output(vdev, vq);
>  
>          if (unlikely(vdev->start_on_kick)) {
> -            vdev->started = true;
> -            vdev->start_on_kick = false;
> +            virtio_set_started(vdev, true);
>          }
>      }
>  
> @@ -1557,8 +1554,7 @@ static void virtio_queue_notify_vq(VirtQueue *vq)
>          vq->handle_output(vdev, vq);
>  
>          if (unlikely(vdev->start_on_kick)) {
> -            vdev->started = true;
> -            vdev->start_on_kick = false;
> +            virtio_set_started(vdev, true);
>          }
>      }
>  }
> @@ -1579,8 +1575,7 @@ void virtio_queue_notify(VirtIODevice *vdev, int n)
>      }
>  
>      if (unlikely(vdev->start_on_kick)) {
> -        vdev->started = true;
> -        vdev->start_on_kick = false;
> +        virtio_set_started(vdev, true);
>      }
>  }
>  
> @@ -2291,7 +2286,7 @@ static void virtio_vmstate_change(void *opaque, int running, RunState state)
>      VirtIODevice *vdev = opaque;
>      BusState *qbus = qdev_get_parent_bus(DEVICE(vdev));
>      VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
> -    bool backend_run = running && vdev->started;
> +    bool backend_run = running && virtio_device_started(vdev, vdev->status);
>      vdev->vm_running = running;
>  
>      if (backend_run) {
> @@ -2669,6 +2664,7 @@ static void virtio_device_instance_finalize(Object *obj)
>  
>  static Property virtio_properties[] = {
>      DEFINE_VIRTIO_COMMON_FEATURES(VirtIODevice, host_features),
> +    DEFINE_PROP_BOOL("use-started", VirtIODevice, use_started, true),
>      DEFINE_PROP_END_OF_LIST(),
>  };
>  
> diff --git a/include/hw/virtio/virtio.h b/include/hw/virtio/virtio.h
> index 27c0efc3d0..15d5366939 100644
> --- a/include/hw/virtio/virtio.h
> +++ b/include/hw/virtio/virtio.h
> @@ -105,6 +105,7 @@ struct VirtIODevice
>      uint16_t device_id;
>      bool vm_running;
>      bool broken; /* device in invalid state, needs reset */
> +    bool use_started;
>      bool started;
>      bool start_on_kick; /* virtio 1.0 transitional devices support that */
>      VMChangeStateEntry *vmstate;
> @@ -351,4 +352,24 @@ static inline bool virtio_is_big_endian(VirtIODevice *vdev)
>      /* Devices conforming to VIRTIO 1.0 or later are always LE. */
>      return false;
>  }
> +
> +static inline bool virtio_device_started(VirtIODevice *vdev, uint8_t status)
> +{
> +    if (vdev->use_started) {
> +        return vdev->started;
> +    }
> +
> +    return status & VIRTIO_CONFIG_S_DRIVER_OK;
> +}
> +
> +static inline void virtio_set_started(VirtIODevice *vdev, bool started)
> +{
> +    if (started) {
> +        vdev->start_on_kick = false;
> +    }
> +
> +    if (vdev->use_started) {
> +        vdev->started = started;
> +    }
> +}
>  #endif

Re: [Qemu-devel] [PATCH v3 1/5] virtio: add "use-started" property

[Yongji Xie]

On Fri, 14 Jun 2019 at 19:45, Greg Kurz <groug@kaod.org> wrote:
>
> On Fri, 14 Jun 2019 17:31:17 +0800
> elohimes@gmail.com wrote:
>
> > From: Xie Yongji <xieyongji@baidu.com>
> >
> > In order to avoid migration issues, we introduce a "use-started"
> > property to the base virtio device to indicate whether use
> > "started" flag or not. This property will be true by default and
> > set to false when machine type <= 4.0.1.
> >
> > Suggested-by: Greg Kurz <groug@kaod.org>
> > Signed-off-by: Xie Yongji <xieyongji@baidu.com>
> > ---
> >  hw/block/vhost-user-blk.c  |  4 ++--
> >  hw/core/machine.c          |  8 ++++++--
>
> This patch conflicts with latest upstream changes to hw_compat_4_0_1[].
>
> It seems you need to rebase. Also, I'm still not sure how we're supposed
> to handle hw_compat_4_0_1[] versus hw_compat_4_0[]... nobody commented
> on:
>
> https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg00637.html
> https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg00641.html
>
> Maybe worth to sort that out before re-posting.
>

If hw_compat_4_0_1[] is introduced only for q35, I think this patch
should be OK. If not, maybe we should handle hw_compat_4_0_1[] in
other machine types (i440fx, arm, ppc, s390)?

Hi Alex and Paolo,

Any comment for this?

Thanks,
Yongji

Re: [Qemu-devel] [PATCH v3 1/5] virtio: add "use-started" property

[Greg Kurz]

On Mon, 17 Jun 2019 10:14:30 +0800
Yongji Xie <elohimes@gmail.com> wrote:

> On Fri, 14 Jun 2019 at 19:45, Greg Kurz <groug@kaod.org> wrote:
> >
> > On Fri, 14 Jun 2019 17:31:17 +0800
> > elohimes@gmail.com wrote:
> >  
> > > From: Xie Yongji <xieyongji@baidu.com>
> > >
> > > In order to avoid migration issues, we introduce a "use-started"
> > > property to the base virtio device to indicate whether use
> > > "started" flag or not. This property will be true by default and
> > > set to false when machine type <= 4.0.1.
> > >
> > > Suggested-by: Greg Kurz <groug@kaod.org>
> > > Signed-off-by: Xie Yongji <xieyongji@baidu.com>
> > > ---
> > >  hw/block/vhost-user-blk.c  |  4 ++--
> > >  hw/core/machine.c          |  8 ++++++--  
> >
> > This patch conflicts with latest upstream changes to hw_compat_4_0_1[].
> >
> > It seems you need to rebase. Also, I'm still not sure how we're supposed
> > to handle hw_compat_4_0_1[] versus hw_compat_4_0[]... nobody commented
> > on:
> >
> > https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg00637.html
> > https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg00641.html
> >
> > Maybe worth to sort that out before re-posting.
> >  
> 
> If hw_compat_4_0_1[] is introduced only for q35, I think this patch
> should be OK. If not, maybe we should handle hw_compat_4_0_1[] in
> other machine types (i440fx, arm, ppc, s390)?
> 

It turns out that hw_compat_4_0_1[] isn't needed at all. Please see:

https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg03054.html

> Hi Alex and Paolo,
> 
> Any comment for this?
> 
> Thanks,
> Yongji

Re: [Qemu-devel] [PATCH v3 1/5] virtio: add "use-started" property

[Yongji Xie]

On Mon, 17 Jun 2019 at 13:24, Greg Kurz <groug@kaod.org> wrote:
>
> On Mon, 17 Jun 2019 10:14:30 +0800
> Yongji Xie <elohimes@gmail.com> wrote:
>
> > On Fri, 14 Jun 2019 at 19:45, Greg Kurz <groug@kaod.org> wrote:
> > >
> > > On Fri, 14 Jun 2019 17:31:17 +0800
> > > elohimes@gmail.com wrote:
> > >
> > > > From: Xie Yongji <xieyongji@baidu.com>
> > > >
> > > > In order to avoid migration issues, we introduce a "use-started"
> > > > property to the base virtio device to indicate whether use
> > > > "started" flag or not. This property will be true by default and
> > > > set to false when machine type <= 4.0.1.
> > > >
> > > > Suggested-by: Greg Kurz <groug@kaod.org>
> > > > Signed-off-by: Xie Yongji <xieyongji@baidu.com>
> > > > ---
> > > >  hw/block/vhost-user-blk.c  |  4 ++--
> > > >  hw/core/machine.c          |  8 ++++++--
> > >
> > > This patch conflicts with latest upstream changes to hw_compat_4_0_1[].
> > >
> > > It seems you need to rebase. Also, I'm still not sure how we're supposed
> > > to handle hw_compat_4_0_1[] versus hw_compat_4_0[]... nobody commented
> > > on:
> > >
> > > https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg00637.html
> > > https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg00641.html
> > >
> > > Maybe worth to sort that out before re-posting.
> > >
> >
> > If hw_compat_4_0_1[] is introduced only for q35, I think this patch
> > should be OK. If not, maybe we should handle hw_compat_4_0_1[] in
> > other machine types (i440fx, arm, ppc, s390)?
> >
>
> It turns out that hw_compat_4_0_1[] isn't needed at all. Please see:
>
> https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg03054.html
>

Oh, great! I will rebase my patch after this commit is merged.

Thanks,
Yongji

Re: [Qemu-devel] [PATCH v3 1/5] virtio: add "use-started" property

[Greg Kurz]

On Mon, 17 Jun 2019 14:04:10 +0800
Yongji Xie <elohimes@gmail.com> wrote:

> On Mon, 17 Jun 2019 at 13:24, Greg Kurz <groug@kaod.org> wrote:
> >
> > On Mon, 17 Jun 2019 10:14:30 +0800
> > Yongji Xie <elohimes@gmail.com> wrote:
> >  
> > > On Fri, 14 Jun 2019 at 19:45, Greg Kurz <groug@kaod.org> wrote:  
> > > >
> > > > On Fri, 14 Jun 2019 17:31:17 +0800
> > > > elohimes@gmail.com wrote:
> > > >  
> > > > > From: Xie Yongji <xieyongji@baidu.com>
> > > > >
> > > > > In order to avoid migration issues, we introduce a "use-started"
> > > > > property to the base virtio device to indicate whether use
> > > > > "started" flag or not. This property will be true by default and
> > > > > set to false when machine type <= 4.0.1.
> > > > >
> > > > > Suggested-by: Greg Kurz <groug@kaod.org>
> > > > > Signed-off-by: Xie Yongji <xieyongji@baidu.com>
> > > > > ---
> > > > >  hw/block/vhost-user-blk.c  |  4 ++--
> > > > >  hw/core/machine.c          |  8 ++++++--  
> > > >
> > > > This patch conflicts with latest upstream changes to hw_compat_4_0_1[].
> > > >
> > > > It seems you need to rebase. Also, I'm still not sure how we're supposed
> > > > to handle hw_compat_4_0_1[] versus hw_compat_4_0[]... nobody commented
> > > > on:
> > > >
> > > > https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg00637.html
> > > > https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg00641.html
> > > >
> > > > Maybe worth to sort that out before re-posting.
> > > >  
> > >
> > > If hw_compat_4_0_1[] is introduced only for q35, I think this patch
> > > should be OK. If not, maybe we should handle hw_compat_4_0_1[] in
> > > other machine types (i440fx, arm, ppc, s390)?
> > >  
> >
> > It turns out that hw_compat_4_0_1[] isn't needed at all. Please see:
> >
> > https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg03054.html
> >  
> 
> Oh, great! I will rebase my patch after this commit is merged.
> 
> Thanks,
> Yongji

You can proceed.

https://git.qemu.org/?p=qemu.git;a=commit;h=8e8cbed09ad9d577955691b4c061b61b602406d1

Cheers,

--
Greg

Re: [Qemu-devel] [PATCH v3 1/5] virtio: add "use-started" property

[Yongji Xie]

On Sat, 22 Jun 2019 at 23:51, Greg Kurz <groug@kaod.org> wrote:
>
> On Mon, 17 Jun 2019 14:04:10 +0800
> Yongji Xie <elohimes@gmail.com> wrote:
>
> > On Mon, 17 Jun 2019 at 13:24, Greg Kurz <groug@kaod.org> wrote:
> > >
> > > On Mon, 17 Jun 2019 10:14:30 +0800
> > > Yongji Xie <elohimes@gmail.com> wrote:
> > >
> > > > On Fri, 14 Jun 2019 at 19:45, Greg Kurz <groug@kaod.org> wrote:
> > > > >
> > > > > On Fri, 14 Jun 2019 17:31:17 +0800
> > > > > elohimes@gmail.com wrote:
> > > > >
> > > > > > From: Xie Yongji <xieyongji@baidu.com>
> > > > > >
> > > > > > In order to avoid migration issues, we introduce a "use-started"
> > > > > > property to the base virtio device to indicate whether use
> > > > > > "started" flag or not. This property will be true by default and
> > > > > > set to false when machine type <= 4.0.1.
> > > > > >
> > > > > > Suggested-by: Greg Kurz <groug@kaod.org>
> > > > > > Signed-off-by: Xie Yongji <xieyongji@baidu.com>
> > > > > > ---
> > > > > >  hw/block/vhost-user-blk.c  |  4 ++--
> > > > > >  hw/core/machine.c          |  8 ++++++--
> > > > >
> > > > > This patch conflicts with latest upstream changes to hw_compat_4_0_1[].
> > > > >
> > > > > It seems you need to rebase. Also, I'm still not sure how we're supposed
> > > > > to handle hw_compat_4_0_1[] versus hw_compat_4_0[]... nobody commented
> > > > > on:
> > > > >
> > > > > https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg00637.html
> > > > > https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg00641.html
> > > > >
> > > > > Maybe worth to sort that out before re-posting.
> > > > >
> > > >
> > > > If hw_compat_4_0_1[] is introduced only for q35, I think this patch
> > > > should be OK. If not, maybe we should handle hw_compat_4_0_1[] in
> > > > other machine types (i440fx, arm, ppc, s390)?
> > > >
> > >
> > > It turns out that hw_compat_4_0_1[] isn't needed at all. Please see:
> > >
> > > https://lists.gnu.org/archive/html/qemu-devel/2019-06/msg03054.html
> > >
> >
> > Oh, great! I will rebase my patch after this commit is merged.
> >
> > Thanks,
> > Yongji
>
> You can proceed.
>
> https://git.qemu.org/?p=qemu.git;a=commit;h=8e8cbed09ad9d577955691b4c061b61b602406d1
>

OK, Thank you.

Thanks,
Yongji

[Qemu-devel] [PATCH v3 2/5] virtio: Set "start_on_kick" for legacy devices

[elohimes]

From: Xie Yongji <xieyongji@baidu.com>

Besides virtio 1.0 transitional devices, we should also
set "start_on_kick" flag for legacy devices (virtio 0.9).

Signed-off-by: Xie Yongji <xieyongji@baidu.com>
---
 hw/virtio/virtio.c         | 6 ++----
 include/hw/virtio/virtio.h | 2 +-
 2 files changed, 3 insertions(+), 5 deletions(-)

diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 19062fbb96..473881e9ec 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -1212,8 +1212,7 @@ void virtio_reset(void *opaque)
         k->reset(vdev);
     }
 
-    vdev->start_on_kick = (virtio_host_has_feature(vdev, VIRTIO_F_VERSION_1) &&
-                          !virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1));
+    vdev->start_on_kick = !virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1);
     vdev->started = false;
     vdev->broken = false;
     vdev->guest_features = 0;
@@ -2325,8 +2324,7 @@ void virtio_init(VirtIODevice *vdev, const char *name,
             g_malloc0(sizeof(*vdev->vector_queues) * nvectors);
     }
 
-    vdev->start_on_kick = (virtio_host_has_feature(vdev, VIRTIO_F_VERSION_1) &&
-                          !virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1));
+    vdev->start_on_kick = !virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1);
     vdev->started = false;
     vdev->device_id = device_id;
     vdev->status = 0;
diff --git a/include/hw/virtio/virtio.h b/include/hw/virtio/virtio.h
index 15d5366939..b189788cb2 100644
--- a/include/hw/virtio/virtio.h
+++ b/include/hw/virtio/virtio.h
@@ -107,7 +107,7 @@ struct VirtIODevice
     bool broken; /* device in invalid state, needs reset */
     bool use_started;
     bool started;
-    bool start_on_kick; /* virtio 1.0 transitional devices support that */
+    bool start_on_kick; /* when virtio 1.0 feature has not been negotiated */
     VMChangeStateEntry *vmstate;
     char *bus_name;
     uint8_t device_endian;
-- 
2.17.1

Re: [Qemu-devel] [PATCH v3 2/5] virtio: Set "start_on_kick" for legacy devices

[Greg Kurz]

On Fri, 14 Jun 2019 17:31:18 +0800
elohimes@gmail.com wrote:

> From: Xie Yongji <xieyongji@baidu.com>
> 
> Besides virtio 1.0 transitional devices, we should also
> set "start_on_kick" flag for legacy devices (virtio 0.9).
> 
> Signed-off-by: Xie Yongji <xieyongji@baidu.com>
> ---

Reviewed-by: Greg Kurz <groug@kaod.org>

>  hw/virtio/virtio.c         | 6 ++----
>  include/hw/virtio/virtio.h | 2 +-
>  2 files changed, 3 insertions(+), 5 deletions(-)
> 
> diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
> index 19062fbb96..473881e9ec 100644
> --- a/hw/virtio/virtio.c
> +++ b/hw/virtio/virtio.c
> @@ -1212,8 +1212,7 @@ void virtio_reset(void *opaque)
>          k->reset(vdev);
>      }
>  
> -    vdev->start_on_kick = (virtio_host_has_feature(vdev, VIRTIO_F_VERSION_1) &&
> -                          !virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1));
> +    vdev->start_on_kick = !virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1);
>      vdev->started = false;
>      vdev->broken = false;
>      vdev->guest_features = 0;
> @@ -2325,8 +2324,7 @@ void virtio_init(VirtIODevice *vdev, const char *name,
>              g_malloc0(sizeof(*vdev->vector_queues) * nvectors);
>      }
>  
> -    vdev->start_on_kick = (virtio_host_has_feature(vdev, VIRTIO_F_VERSION_1) &&
> -                          !virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1));
> +    vdev->start_on_kick = !virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1);
>      vdev->started = false;
>      vdev->device_id = device_id;
>      vdev->status = 0;
> diff --git a/include/hw/virtio/virtio.h b/include/hw/virtio/virtio.h
> index 15d5366939..b189788cb2 100644
> --- a/include/hw/virtio/virtio.h
> +++ b/include/hw/virtio/virtio.h
> @@ -107,7 +107,7 @@ struct VirtIODevice
>      bool broken; /* device in invalid state, needs reset */
>      bool use_started;
>      bool started;
> -    bool start_on_kick; /* virtio 1.0 transitional devices support that */
> +    bool start_on_kick; /* when virtio 1.0 feature has not been negotiated */
>      VMChangeStateEntry *vmstate;
>      char *bus_name;
>      uint8_t device_endian;

[Qemu-devel] [PATCH v3 3/5] virtio: Set "start_on_kick" on virtio_set_features()

[elohimes]

From: Xie Yongji <xieyongji@baidu.com>

The guest feature is not set correctly on virtio_reset() and
virtio_init(). So we should not use it to set "start_on_kick" at that
point. This patch set "start_on_kick" on virtio_set_features() instead.

Fixes: badaf79cfdbd3 ("virtio: Introduce started flag to VirtioDevice")
Signed-off-by: Xie Yongji <xieyongji@baidu.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
---
 hw/virtio/virtio.c | 28 ++++++++++++++++++++--------
 1 file changed, 20 insertions(+), 8 deletions(-)

diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 473881e9ec..034320d277 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -1212,7 +1212,7 @@ void virtio_reset(void *opaque)
         k->reset(vdev);
     }
 
-    vdev->start_on_kick = !virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1);
+    vdev->start_on_kick = false;
     vdev->started = false;
     vdev->broken = false;
     vdev->guest_features = 0;
@@ -2063,14 +2063,21 @@ int virtio_set_features(VirtIODevice *vdev, uint64_t val)
         return -EINVAL;
     }
     ret = virtio_set_features_nocheck(vdev, val);
-    if (!ret && virtio_vdev_has_feature(vdev, VIRTIO_RING_F_EVENT_IDX)) {
-        /* VIRTIO_RING_F_EVENT_IDX changes the size of the caches.  */
-        int i;
-        for (i = 0; i < VIRTIO_QUEUE_MAX; i++) {
-            if (vdev->vq[i].vring.num != 0) {
-                virtio_init_region_cache(vdev, i);
+    if (!ret) {
+        if (virtio_vdev_has_feature(vdev, VIRTIO_RING_F_EVENT_IDX)) {
+            /* VIRTIO_RING_F_EVENT_IDX changes the size of the caches.  */
+            int i;
+            for (i = 0; i < VIRTIO_QUEUE_MAX; i++) {
+                if (vdev->vq[i].vring.num != 0) {
+                    virtio_init_region_cache(vdev, i);
+                }
             }
         }
+
+        if (!virtio_device_started(vdev, vdev->status) &&
+            !virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1)) {
+            vdev->start_on_kick = true;
+        }
     }
     return ret;
 }
@@ -2222,6 +2229,11 @@ int virtio_load(VirtIODevice *vdev, QEMUFile *f, int version_id)
         }
     }
 
+    if (!virtio_device_started(vdev, vdev->status) &&
+        !virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1)) {
+        vdev->start_on_kick = true;
+    }
+
     rcu_read_lock();
     for (i = 0; i < num; i++) {
         if (vdev->vq[i].vring.desc) {
@@ -2324,7 +2336,7 @@ void virtio_init(VirtIODevice *vdev, const char *name,
             g_malloc0(sizeof(*vdev->vector_queues) * nvectors);
     }
 
-    vdev->start_on_kick = !virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1);
+    vdev->start_on_kick = false;
     vdev->started = false;
     vdev->device_id = device_id;
     vdev->status = 0;
-- 
2.17.1

[Qemu-devel] [PATCH v3 4/5] virtio: Make sure we get correct state of device on handle_aio_output()

[elohimes]

From: Xie Yongji <xieyongji@baidu.com>

We should set the flags: "start_on_kick" and "started" after we call
the kick functions (handle_aio_output() and handle_output()).

Signed-off-by: Xie Yongji <xieyongji@baidu.com>
---
 hw/virtio/virtio.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 034320d277..b4301f9e02 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -1571,10 +1571,10 @@ void virtio_queue_notify(VirtIODevice *vdev, int n)
         event_notifier_set(&vq->host_notifier);
     } else if (vq->handle_output) {
         vq->handle_output(vdev, vq);
-    }
 
-    if (unlikely(vdev->start_on_kick)) {
-        virtio_set_started(vdev, true);
+        if (unlikely(vdev->start_on_kick)) {
+            virtio_set_started(vdev, true);
+        }
     }
 }
 
-- 
2.17.1

[Qemu-devel] [PATCH v3 5/5] virtio: Don't change "started" flag on virtio_vmstate_change()

[elohimes]

From: Xie Yongji <xieyongji@baidu.com>

We will call virtio_set_status() on virtio_vmstate_change().
The "started" flag should not be changed in this case. Otherwise,
we may get an incorrect value when we set "started" flag but
not set DRIVER_OK in source VM.

Signed-off-by: Xie Yongji <xieyongji@baidu.com>
---
 hw/virtio/virtio.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index b4301f9e02..9af2e339af 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -1163,7 +1163,10 @@ int virtio_set_status(VirtIODevice *vdev, uint8_t val)
         }
     }
 
-    virtio_set_started(vdev, val & VIRTIO_CONFIG_S_DRIVER_OK);
+    if ((vdev->status & VIRTIO_CONFIG_S_DRIVER_OK) !=
+        (val & VIRTIO_CONFIG_S_DRIVER_OK)) {
+        virtio_set_started(vdev, val & VIRTIO_CONFIG_S_DRIVER_OK);
+    }
 
     if (k->set_status) {
         k->set_status(vdev, val);
-- 
2.17.1