[PATCH v4 1/2] Revert "vhost-user: fix lost reconnect"
This reverts commit f02a4b8e6431598612466f76aac64ab492849abf.
Since the current patch cannot completely fix the lost reconnect
problem, there is a scenario that is not considered:
- When the virtio-blk driver is removed from the guest os,
s->connected has no chance to be set to false, resulting in
subsequent reconnection not being executed.
The next patch will completely fix this issue with a better approach.
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 2 +-
hw/scsi/vhost-user-scsi.c | 3 +--
hw/virtio/vhost-user-base.c| 2 +-
hw/virtio/vhost-user.c | 10 ++
include/hw/virtio/vhost-user.h | 3 +--
5 files changed, 6 insertions(+), 14 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 9e6bbc6950..41d1ac3a5a 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -384,7 +384,7 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent
event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >dev,
- vhost_user_blk_disconnect,
vhost_user_blk_event);
+ vhost_user_blk_disconnect);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index a63b1f4948..48a59e020e 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -214,8 +214,7 @@ static void vhost_user_scsi_event(void *opaque,
QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >conf.chardev, >dev,
- vhost_user_scsi_disconnect,
- vhost_user_scsi_event);
+ vhost_user_scsi_disconnect);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user-base.c b/hw/virtio/vhost-user-base.c
index a83167191e..4b54255682 100644
--- a/hw/virtio/vhost-user-base.c
+++ b/hw/virtio/vhost-user-base.c
@@ -254,7 +254,7 @@ static void vub_event(void *opaque, QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >vhost_dev,
- vub_disconnect, vub_event);
+ vub_disconnect);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index cdf9af4a4b..c929097e87 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2776,7 +2776,6 @@ typedef struct {
DeviceState *dev;
CharBackend *cd;
struct vhost_dev *vhost;
-IOEventHandler *event_cb;
} VhostAsyncCallback;
static void vhost_user_async_close_bh(void *opaque)
@@ -2791,10 +2790,7 @@ static void vhost_user_async_close_bh(void *opaque)
*/
if (vhost->vdev) {
data->cb(data->dev);
-} else if (data->event_cb) {
-qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
- NULL, data->dev, NULL, true);
- }
+}
g_free(data);
}
@@ -2806,8 +2802,7 @@ static void vhost_user_async_close_bh(void *opaque)
*/
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb,
-IOEventHandler *event_cb)
+vu_async_close_fn cb)
{
if (!runstate_check(RUN_STATE_SHUTDOWN)) {
/*
@@ -2823,7 +2818,6 @@ void vhost_user_async_close(DeviceState *d,
data->dev = d;
data->cd = chardev;
data->vhost = vhost;
-data->event_cb = event_cb;
/* Disable any further notifications on the chardev */
qemu_chr_fe_set_handlers(chardev,
diff --git a/include/hw/virtio/vhost-user.h b/include/hw/virtio/vhost-user.h
index d7c09ffd34..324cd8663a 100644
--- a/include/hw/virtio/vhost-user.h
+++ b/include/hw/virtio/vhost-user.h
@@ -108,7 +108,6 @@ typedef void (*vu_async_close_fn)(DeviceState *cb);
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb,
-IOEventHandler *event_cb);
+vu_async_close_fn cb);
#endif
--
2.45.0
[PATCH v4 2/2] vhost-user: fix lost reconnect again
When the vhost-user is reconnecting to the backend, and if the vhost-user fails
at the get_features in vhost_dev_init(), then the reconnect will fail
and it will not be retriggered forever.
The reason is:
When the vhost-user fail at get_features, the vhost_dev_cleanup will be called
immediately.
vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
The reconnect path is:
vhost_user_blk_event
vhost_user_async_close(.. vhost_user_blk_disconnect ..)
qemu_chr_fe_set_handlers <- clear the notifier callback
schedule vhost_user_async_close_bh
The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
called, then the event fd callback will not be reinstalled.
We need to ensure that even if vhost_dev_init initialization fails, the event
handler still needs to be reinstalled when s->connected is false.
All vhost-user devices have this issue, including vhost-user-blk/scsi.
Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 3 ++-
hw/scsi/vhost-user-scsi.c | 3 ++-
hw/virtio/vhost-user-base.c | 3 ++-
hw/virtio/vhost-user.c | 10 +-
4 files changed, 7 insertions(+), 12 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 41d1ac3a5a..c6842ced48 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -353,7 +353,7 @@ static void vhost_user_blk_disconnect(DeviceState *dev)
VHostUserBlk *s = VHOST_USER_BLK(vdev);
if (!s->connected) {
-return;
+goto done;
}
s->connected = false;
@@ -361,6 +361,7 @@ static void vhost_user_blk_disconnect(DeviceState *dev)
vhost_dev_cleanup(>dev);
+done:
/* Re-instate the event handler for new connections */
qemu_chr_fe_set_handlers(>chardev, NULL, NULL, vhost_user_blk_event,
NULL, dev, NULL, true);
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 48a59e020e..b49a11d23b 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -181,7 +181,7 @@ static void vhost_user_scsi_disconnect(DeviceState *dev)
VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
if (!s->connected) {
-return;
+goto done;
}
s->connected = false;
@@ -189,6 +189,7 @@ static void vhost_user_scsi_disconnect(DeviceState *dev)
vhost_dev_cleanup(>dev);
+done:
/* Re-instate the event handler for new connections */
qemu_chr_fe_set_handlers(>conf.chardev, NULL, NULL,
vhost_user_scsi_event, NULL, dev, NULL, true);
diff --git a/hw/virtio/vhost-user-base.c b/hw/virtio/vhost-user-base.c
index 4b54255682..11e72b1e3b 100644
--- a/hw/virtio/vhost-user-base.c
+++ b/hw/virtio/vhost-user-base.c
@@ -225,13 +225,14 @@ static void vub_disconnect(DeviceState *dev)
VHostUserBase *vub = VHOST_USER_BASE(vdev);
if (!vub->connected) {
-return;
+goto done;
}
vub->connected = false;
vub_stop(vdev);
vhost_dev_cleanup(>vhost_dev);
+done:
/* Re-instate the event handler for new connections */
qemu_chr_fe_set_handlers(>chardev,
NULL, NULL, vub_event,
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index c929097e87..c407ea8939 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2781,16 +2781,8 @@ typedef struct {
static void vhost_user_async_close_bh(void *opaque)
{
VhostAsyncCallback *data = opaque;
-struct vhost_dev *vhost = data->vhost;
-/*
- * If the vhost_dev has been cleared in the meantime there is
- * nothing left to do as some other path has completed the
- * cleanup.
- */
-if (vhost->vdev) {
-data->cb(data->dev);
-}
+data->cb(data->dev);
g_free(data);
}
--
2.45.0
Re: [PATCH v3 2/2] vhost-user: fix lost reconnect again
> 2024年5月15日 23:47，Raphael Norwitz 写道：
>
> The case your describing makes sense but now I have some concerns on
> the vhost_dev_cleanup bit.
>
> On Wed, May 15, 2024 at 1:47 AM Li Feng <mailto:fen...@smartx.com>> wrote:
>>
>>
>>
>>> 2024年5月14日 21:58，Raphael Norwitz 写道：
>>>
>>> Code looks good. Just a question on the error case you're trying to fix.
>>>
>>> On Tue, May 14, 2024 at 2:12 AM Li Feng wrote:
>>>>
>>>> When the vhost-user is reconnecting to the backend, and if the vhost-user
>>>> fails
>>>> at the get_features in vhost_dev_init(), then the reconnect will fail
>>>> and it will not be retriggered forever.
>>>>
>>>> The reason is:
>>>> When the vhost-user fail at get_features, the vhost_dev_cleanup will be
>>>> called
>>>> immediately.
>>>>
>>>> vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
>>>>
>>>> The reconnect path is:
>>>> vhost_user_blk_event
>>>> vhost_user_async_close(.. vhost_user_blk_disconnect ..)
>>>>qemu_chr_fe_set_handlers <- clear the notifier callback
>>>> schedule vhost_user_async_close_bh
>>>>
>>>> The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
>>>> called, then the event fd callback will not be reinstalled.
>>>>
>>>> With this patch, the vhost_user_blk_disconnect will call the
>>>> vhost_dev_cleanup() again, it's safe.
>>>>
>>>> In addition, the CLOSE event may occur in a scenario where connected is
>>>> false.
>>>> At this time, the event handler will be cleared. We need to ensure that the
>>>> event handler can remain installed.
>>>
>>> Following on from the prior patch, why would "connected" be false when
>>> a CLOSE event happens?
>>
>> In OPEN event handling, vhost_user_blk_connect calls vhost_dev_init and
>> encounters
>> an error such that s->connected remains false.
>> Next, after the CLOSE event arrives, it is found that s->connected is false,
>> so nothing
>> is done, but the event handler will be cleaned up in `vhost_user_async_close`
>> before the CLOSE event is executed.
>>
>
> Got it - I see why the event handler is never re-installed in the code
> as it was before if we fail at get_features. That said, how do you
> explain your comment:
OK, I will update the commit message because this code has changed some months
ago.
>
>>>> With this patch, the vhost_user_blk_disconnect will call the
>>>> vhost_dev_cleanup() again, it's safe.
>
> I see vhost_dev_cleanup() accessing hdev without even a NULL check. In
> the case we're talking about here I don't think it's a problem because
> if vhost_dev_init() fails, connected will be false and hit the goto
> but I am concerned that there could be double-frees or use-after-frees
> in other cases.
OK, you are right, with this patch, the vhost_dev_cleanup will not be
called multiple times now.
I think there is no need to worry about calling vhost_dev_cleanup multiple
times,
because historically vhost_dev_cleanup has been allowed to be called multiple
times, and looking at the code, it can be found that calling vhost_dev_cleanup
multiple times is indeed safe.
Look this patch:
commit e0547b59dc0ead4c605d3f02d1c8829630a1311b
Author: Marc-André Lureau
Date: Wed Jul 27 01:15:02 2016 +0400
vhost: make vhost_dev_cleanup() idempotent
It is called on multiple code path, so make it safe to call several
times (note: I don't remember a reproducer here, but a function called
'cleanup' should probably be idempotent in my book)
Signed-off-by: Marc-André Lureau
Reviewed-by: Michael S. Tsirkin
Signed-off-by: Michael S. Tsirkin
Thanks,
Li
>
>> Thanks,
>> Li
>>
>>>
>>>>
>>>> All vhost-user devices have this issue, including vhost-user-blk/scsi.
>>>>
>>>> Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
>>>>
>>>> Signed-off-by: Li Feng
>>>> ---
>>>> hw/block/vhost-user-blk.c | 3 ++-
>>>> hw/scsi/vhost-user-scsi.c | 3 ++-
>>>> hw/virtio/vhost-user-base.c | 3 ++-
>>>> hw/virtio/vhost-user.c | 10 +-
>>>> 4 files changed, 7 insertions(+), 12 deletions(-)
>>>>
>>>> diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
>>>> index 41d1ac3a5a..c6842ced48 1006
Re: [PATCH v3 2/2] vhost-user: fix lost reconnect again
> 2024年5月14日 21:58，Raphael Norwitz 写道：
>
> Code looks good. Just a question on the error case you're trying to fix.
>
> On Tue, May 14, 2024 at 2:12 AM Li Feng wrote:
>>
>> When the vhost-user is reconnecting to the backend, and if the vhost-user
>> fails
>> at the get_features in vhost_dev_init(), then the reconnect will fail
>> and it will not be retriggered forever.
>>
>> The reason is:
>> When the vhost-user fail at get_features, the vhost_dev_cleanup will be
>> called
>> immediately.
>>
>> vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
>>
>> The reconnect path is:
>> vhost_user_blk_event
>> vhost_user_async_close(.. vhost_user_blk_disconnect ..)
>> qemu_chr_fe_set_handlers <- clear the notifier callback
>> schedule vhost_user_async_close_bh
>>
>> The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
>> called, then the event fd callback will not be reinstalled.
>>
>> With this patch, the vhost_user_blk_disconnect will call the
>> vhost_dev_cleanup() again, it's safe.
>>
>> In addition, the CLOSE event may occur in a scenario where connected is
>> false.
>> At this time, the event handler will be cleared. We need to ensure that the
>> event handler can remain installed.
>
> Following on from the prior patch, why would "connected" be false when
> a CLOSE event happens?
In OPEN event handling, vhost_user_blk_connect calls vhost_dev_init and
encounters
an error such that s->connected remains false.
Next, after the CLOSE event arrives, it is found that s->connected is false, so
nothing
is done, but the event handler will be cleaned up in `vhost_user_async_close`
before the CLOSE event is executed.
Thanks,
Li
>
>>
>> All vhost-user devices have this issue, including vhost-user-blk/scsi.
>>
>> Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
>>
>> Signed-off-by: Li Feng
>> ---
>> hw/block/vhost-user-blk.c | 3 ++-
>> hw/scsi/vhost-user-scsi.c | 3 ++-
>> hw/virtio/vhost-user-base.c | 3 ++-
>> hw/virtio/vhost-user.c | 10 +-
>> 4 files changed, 7 insertions(+), 12 deletions(-)
>>
>> diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
>> index 41d1ac3a5a..c6842ced48 100644
>> --- a/hw/block/vhost-user-blk.c
>> +++ b/hw/block/vhost-user-blk.c
>> @@ -353,7 +353,7 @@ static void vhost_user_blk_disconnect(DeviceState *dev)
>> VHostUserBlk *s = VHOST_USER_BLK(vdev);
>>
>> if (!s->connected) {
>> -return;
>> +goto done;
>> }
>> s->connected = false;
>>
>> @@ -361,6 +361,7 @@ static void vhost_user_blk_disconnect(DeviceState *dev)
>>
>> vhost_dev_cleanup(>dev);
>>
>> +done:
>> /* Re-instate the event handler for new connections */
>> qemu_chr_fe_set_handlers(>chardev, NULL, NULL, vhost_user_blk_event,
>> NULL, dev, NULL, true);
>> diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
>> index 48a59e020e..b49a11d23b 100644
>> --- a/hw/scsi/vhost-user-scsi.c
>> +++ b/hw/scsi/vhost-user-scsi.c
>> @@ -181,7 +181,7 @@ static void vhost_user_scsi_disconnect(DeviceState *dev)
>> VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
>>
>> if (!s->connected) {
>> -return;
>> +goto done;
>> }
>> s->connected = false;
>>
>> @@ -189,6 +189,7 @@ static void vhost_user_scsi_disconnect(DeviceState *dev)
>>
>> vhost_dev_cleanup(>dev);
>>
>> +done:
>> /* Re-instate the event handler for new connections */
>> qemu_chr_fe_set_handlers(>conf.chardev, NULL, NULL,
>> vhost_user_scsi_event, NULL, dev, NULL, true);
>> diff --git a/hw/virtio/vhost-user-base.c b/hw/virtio/vhost-user-base.c
>> index 4b54255682..11e72b1e3b 100644
>> --- a/hw/virtio/vhost-user-base.c
>> +++ b/hw/virtio/vhost-user-base.c
>> @@ -225,13 +225,14 @@ static void vub_disconnect(DeviceState *dev)
>> VHostUserBase *vub = VHOST_USER_BASE(vdev);
>>
>> if (!vub->connected) {
>> -return;
>> +goto done;
>> }
>> vub->connected = false;
>>
>> vub_stop(vdev);
>> vhost_dev_cleanup(>vhost_dev);
>>
>> +done:
>> /* Re-instate the event handler for new connections */
>> qemu_chr_fe_set_handlers(>chardev,
>> NULL, NULL, vub_event,
>> diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
>> index c929097e87..c407ea8939 100644
>> --- a/hw/virtio/vhost-user.c
>> +++ b/hw/virtio/vhost-user.c
>> @@ -2781,16 +2781,8 @@ typedef struct {
>> static void vhost_user_async_close_bh(void *opaque)
>> {
>> VhostAsyncCallback *data = opaque;
>> -struct vhost_dev *vhost = data->vhost;
>>
>> -/*
>> - * If the vhost_dev has been cleared in the meantime there is
>> - * nothing left to do as some other path has completed the
>> - * cleanup.
>> - */
>> -if (vhost->vdev) {
>> -data->cb(data->dev);
>> -}
>> +data->cb(data->dev);
>>
>> g_free(data);
>> }
>> --
>> 2.45.0
>>
Re: [PATCH v3 1/2] Revert "vhost-user: fix lost reconnect"
> 2024年5月14日 21:58，Raphael Norwitz 写道：
>
> The code for these two patches looks fine. Just some questions on the
> failure case you're trying to fix.
>
>
> On Tue, May 14, 2024 at 2:12 AM Li Feng wrote:
>>
>> This reverts commit f02a4b8e6431598612466f76aac64ab492849abf.
>>
>> Since the current patch cannot completely fix the lost reconnect
>> problem, there is a scenario that is not considered:
>> - When the virtio-blk driver is removed from the guest os,
>> s->connected has no chance to be set to false, resulting in
>
> Why would the virtio-blk driver being removed (unloaded?) in the guest
> effect s->connected? Isn't this variable just tracking whether Qemu is
> connected to the backend process? What does it have to do with the
> guest driver state?
Unload the virtio-blk, it will trigger ‘vhost_user_blk_stop’, and in
`vhost_dev_stop`
it will set the `hdev->vdev = NULL;`.
Next if kill the backend, the CLOSE event will be triggered, and the
`vhost->vdev`
has been set to null before, then the `vhost_user_blk_disconnect` will not have
a
chance to execute.So that he s->connected is still true.
static void vhost_user_async_close_bh(void *opaque)
{
VhostAsyncCallback *data = opaque;
struct vhost_dev *vhost = data->vhost;
/*
* If the vhost_dev has been cleared in the meantime there is
* nothing left to do as some other path has completed the
* cleanup.
*/
if (vhost->vdev) { < HERE vdev is null.
data->cb(data->dev);
} else if (data->event_cb) {
qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
NULL, data->dev, NULL, true);
}
g_free(data);
}
Thanks,
Li
>
>> subsequent reconnection not being executed.
>>
>> The next patch will completely fix this issue with a better approach.
>>
>> Signed-off-by: Li Feng
>> ---
>> hw/block/vhost-user-blk.c | 2 +-
>> hw/scsi/vhost-user-scsi.c | 3 +--
>> hw/virtio/vhost-user-base.c| 2 +-
>> hw/virtio/vhost-user.c | 10 ++
>> include/hw/virtio/vhost-user.h | 3 +--
>> 5 files changed, 6 insertions(+), 14 deletions(-)
>>
>> diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
>> index 9e6bbc6950..41d1ac3a5a 100644
>> --- a/hw/block/vhost-user-blk.c
>> +++ b/hw/block/vhost-user-blk.c
>> @@ -384,7 +384,7 @@ static void vhost_user_blk_event(void *opaque,
>> QEMUChrEvent event)
>> case CHR_EVENT_CLOSED:
>> /* defer close until later to avoid circular close */
>> vhost_user_async_close(dev, >chardev, >dev,
>> - vhost_user_blk_disconnect,
>> vhost_user_blk_event);
>> + vhost_user_blk_disconnect);
>> break;
>> case CHR_EVENT_BREAK:
>> case CHR_EVENT_MUX_IN:
>> diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
>> index a63b1f4948..48a59e020e 100644
>> --- a/hw/scsi/vhost-user-scsi.c
>> +++ b/hw/scsi/vhost-user-scsi.c
>> @@ -214,8 +214,7 @@ static void vhost_user_scsi_event(void *opaque,
>> QEMUChrEvent event)
>> case CHR_EVENT_CLOSED:
>> /* defer close until later to avoid circular close */
>> vhost_user_async_close(dev, >conf.chardev, >dev,
>> - vhost_user_scsi_disconnect,
>> - vhost_user_scsi_event);
>> + vhost_user_scsi_disconnect);
>> break;
>> case CHR_EVENT_BREAK:
>> case CHR_EVENT_MUX_IN:
>> diff --git a/hw/virtio/vhost-user-base.c b/hw/virtio/vhost-user-base.c
>> index a83167191e..4b54255682 100644
>> --- a/hw/virtio/vhost-user-base.c
>> +++ b/hw/virtio/vhost-user-base.c
>> @@ -254,7 +254,7 @@ static void vub_event(void *opaque, QEMUChrEvent event)
>> case CHR_EVENT_CLOSED:
>> /* defer close until later to avoid circular close */
>> vhost_user_async_close(dev, >chardev, >vhost_dev,
>> - vub_disconnect, vub_event);
>> + vub_disconnect);
>> break;
>> case CHR_EVENT_BREAK:
>> case CHR_EVENT_MUX_IN:
>> diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
>> index cdf9af4a4b..c929097e87 100644
>> --- a/hw/virtio/vhost-user.c
>> +++ b/hw/virtio/vhost-user.c
>> @@ -2776,7 +2776,6 @@ typedef struct {
>> DeviceState *dev;
>> CharBackend *cd;
>> struct vhost_dev *vhost;
>> -IOEventHandler *event_cb;
>
[PATCH v3 2/2] vhost-user: fix lost reconnect again
When the vhost-user is reconnecting to the backend, and if the vhost-user fails
at the get_features in vhost_dev_init(), then the reconnect will fail
and it will not be retriggered forever.
The reason is:
When the vhost-user fail at get_features, the vhost_dev_cleanup will be called
immediately.
vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
The reconnect path is:
vhost_user_blk_event
vhost_user_async_close(.. vhost_user_blk_disconnect ..)
qemu_chr_fe_set_handlers <- clear the notifier callback
schedule vhost_user_async_close_bh
The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
called, then the event fd callback will not be reinstalled.
With this patch, the vhost_user_blk_disconnect will call the
vhost_dev_cleanup() again, it's safe.
In addition, the CLOSE event may occur in a scenario where connected is false.
At this time, the event handler will be cleared. We need to ensure that the
event handler can remain installed.
All vhost-user devices have this issue, including vhost-user-blk/scsi.
Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 3 ++-
hw/scsi/vhost-user-scsi.c | 3 ++-
hw/virtio/vhost-user-base.c | 3 ++-
hw/virtio/vhost-user.c | 10 +-
4 files changed, 7 insertions(+), 12 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 41d1ac3a5a..c6842ced48 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -353,7 +353,7 @@ static void vhost_user_blk_disconnect(DeviceState *dev)
VHostUserBlk *s = VHOST_USER_BLK(vdev);
if (!s->connected) {
-return;
+goto done;
}
s->connected = false;
@@ -361,6 +361,7 @@ static void vhost_user_blk_disconnect(DeviceState *dev)
vhost_dev_cleanup(>dev);
+done:
/* Re-instate the event handler for new connections */
qemu_chr_fe_set_handlers(>chardev, NULL, NULL, vhost_user_blk_event,
NULL, dev, NULL, true);
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 48a59e020e..b49a11d23b 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -181,7 +181,7 @@ static void vhost_user_scsi_disconnect(DeviceState *dev)
VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
if (!s->connected) {
-return;
+goto done;
}
s->connected = false;
@@ -189,6 +189,7 @@ static void vhost_user_scsi_disconnect(DeviceState *dev)
vhost_dev_cleanup(>dev);
+done:
/* Re-instate the event handler for new connections */
qemu_chr_fe_set_handlers(>conf.chardev, NULL, NULL,
vhost_user_scsi_event, NULL, dev, NULL, true);
diff --git a/hw/virtio/vhost-user-base.c b/hw/virtio/vhost-user-base.c
index 4b54255682..11e72b1e3b 100644
--- a/hw/virtio/vhost-user-base.c
+++ b/hw/virtio/vhost-user-base.c
@@ -225,13 +225,14 @@ static void vub_disconnect(DeviceState *dev)
VHostUserBase *vub = VHOST_USER_BASE(vdev);
if (!vub->connected) {
-return;
+goto done;
}
vub->connected = false;
vub_stop(vdev);
vhost_dev_cleanup(>vhost_dev);
+done:
/* Re-instate the event handler for new connections */
qemu_chr_fe_set_handlers(>chardev,
NULL, NULL, vub_event,
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index c929097e87..c407ea8939 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2781,16 +2781,8 @@ typedef struct {
static void vhost_user_async_close_bh(void *opaque)
{
VhostAsyncCallback *data = opaque;
-struct vhost_dev *vhost = data->vhost;
-/*
- * If the vhost_dev has been cleared in the meantime there is
- * nothing left to do as some other path has completed the
- * cleanup.
- */
-if (vhost->vdev) {
-data->cb(data->dev);
-}
+data->cb(data->dev);
g_free(data);
}
--
2.45.0
[PATCH v3 1/2] Revert "vhost-user: fix lost reconnect"
This reverts commit f02a4b8e6431598612466f76aac64ab492849abf.
Since the current patch cannot completely fix the lost reconnect
problem, there is a scenario that is not considered:
- When the virtio-blk driver is removed from the guest os,
s->connected has no chance to be set to false, resulting in
subsequent reconnection not being executed.
The next patch will completely fix this issue with a better approach.
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 2 +-
hw/scsi/vhost-user-scsi.c | 3 +--
hw/virtio/vhost-user-base.c| 2 +-
hw/virtio/vhost-user.c | 10 ++
include/hw/virtio/vhost-user.h | 3 +--
5 files changed, 6 insertions(+), 14 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 9e6bbc6950..41d1ac3a5a 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -384,7 +384,7 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent
event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >dev,
- vhost_user_blk_disconnect,
vhost_user_blk_event);
+ vhost_user_blk_disconnect);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index a63b1f4948..48a59e020e 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -214,8 +214,7 @@ static void vhost_user_scsi_event(void *opaque,
QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >conf.chardev, >dev,
- vhost_user_scsi_disconnect,
- vhost_user_scsi_event);
+ vhost_user_scsi_disconnect);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user-base.c b/hw/virtio/vhost-user-base.c
index a83167191e..4b54255682 100644
--- a/hw/virtio/vhost-user-base.c
+++ b/hw/virtio/vhost-user-base.c
@@ -254,7 +254,7 @@ static void vub_event(void *opaque, QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >vhost_dev,
- vub_disconnect, vub_event);
+ vub_disconnect);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index cdf9af4a4b..c929097e87 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2776,7 +2776,6 @@ typedef struct {
DeviceState *dev;
CharBackend *cd;
struct vhost_dev *vhost;
-IOEventHandler *event_cb;
} VhostAsyncCallback;
static void vhost_user_async_close_bh(void *opaque)
@@ -2791,10 +2790,7 @@ static void vhost_user_async_close_bh(void *opaque)
*/
if (vhost->vdev) {
data->cb(data->dev);
-} else if (data->event_cb) {
-qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
- NULL, data->dev, NULL, true);
- }
+}
g_free(data);
}
@@ -2806,8 +2802,7 @@ static void vhost_user_async_close_bh(void *opaque)
*/
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb,
-IOEventHandler *event_cb)
+vu_async_close_fn cb)
{
if (!runstate_check(RUN_STATE_SHUTDOWN)) {
/*
@@ -2823,7 +2818,6 @@ void vhost_user_async_close(DeviceState *d,
data->dev = d;
data->cd = chardev;
data->vhost = vhost;
-data->event_cb = event_cb;
/* Disable any further notifications on the chardev */
qemu_chr_fe_set_handlers(chardev,
diff --git a/include/hw/virtio/vhost-user.h b/include/hw/virtio/vhost-user.h
index d7c09ffd34..324cd8663a 100644
--- a/include/hw/virtio/vhost-user.h
+++ b/include/hw/virtio/vhost-user.h
@@ -108,7 +108,6 @@ typedef void (*vu_async_close_fn)(DeviceState *cb);
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb,
-IOEventHandler *event_cb);
+vu_async_close_fn cb);
#endif
--
2.45.0
[PATCH v2 2/2] vhost-user: fix lost reconnect again
When the vhost-user is reconnecting to the backend, and if the vhost-user fails
at the get_features in vhost_dev_init(), then the reconnect will fail
and it will not be retriggered forever.
The reason is:
When the vhost-user fail at get_features, the vhost_dev_cleanup will be called
immediately.
vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
The reconnect path is:
vhost_user_blk_event
vhost_user_async_close(.. vhost_user_blk_disconnect ..)
qemu_chr_fe_set_handlers <- clear the notifier callback
schedule vhost_user_async_close_bh
The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
called, then the event fd callback will not be reinstalled.
With this patch, the vhost_user_blk_disconnect will call the
vhost_dev_cleanup() again, it's safe.
In addition, the CLOSE event may occur in a scenario where connected is false.
At this time, the event handler will be cleared. We need to ensure that the
event handler can remain installed.
All vhost-user devices have this issue, including vhost-user-blk/scsi.
Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 3 ++-
hw/scsi/vhost-user-scsi.c | 3 ++-
hw/virtio/vhost-user-base.c | 3 ++-
hw/virtio/vhost-user.c | 10 +-
4 files changed, 7 insertions(+), 12 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 41d1ac3a5a..c6842ced48 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -353,7 +353,7 @@ static void vhost_user_blk_disconnect(DeviceState *dev)
VHostUserBlk *s = VHOST_USER_BLK(vdev);
if (!s->connected) {
-return;
+goto done;
}
s->connected = false;
@@ -361,6 +361,7 @@ static void vhost_user_blk_disconnect(DeviceState *dev)
vhost_dev_cleanup(>dev);
+done:
/* Re-instate the event handler for new connections */
qemu_chr_fe_set_handlers(>chardev, NULL, NULL, vhost_user_blk_event,
NULL, dev, NULL, true);
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 48a59e020e..b49a11d23b 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -181,7 +181,7 @@ static void vhost_user_scsi_disconnect(DeviceState *dev)
VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
if (!s->connected) {
-return;
+goto done;
}
s->connected = false;
@@ -189,6 +189,7 @@ static void vhost_user_scsi_disconnect(DeviceState *dev)
vhost_dev_cleanup(>dev);
+done:
/* Re-instate the event handler for new connections */
qemu_chr_fe_set_handlers(>conf.chardev, NULL, NULL,
vhost_user_scsi_event, NULL, dev, NULL, true);
diff --git a/hw/virtio/vhost-user-base.c b/hw/virtio/vhost-user-base.c
index 4b54255682..11e72b1e3b 100644
--- a/hw/virtio/vhost-user-base.c
+++ b/hw/virtio/vhost-user-base.c
@@ -225,13 +225,14 @@ static void vub_disconnect(DeviceState *dev)
VHostUserBase *vub = VHOST_USER_BASE(vdev);
if (!vub->connected) {
-return;
+goto done;
}
vub->connected = false;
vub_stop(vdev);
vhost_dev_cleanup(>vhost_dev);
+done:
/* Re-instate the event handler for new connections */
qemu_chr_fe_set_handlers(>chardev,
NULL, NULL, vub_event,
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index c929097e87..c407ea8939 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2781,16 +2781,8 @@ typedef struct {
static void vhost_user_async_close_bh(void *opaque)
{
VhostAsyncCallback *data = opaque;
-struct vhost_dev *vhost = data->vhost;
-/*
- * If the vhost_dev has been cleared in the meantime there is
- * nothing left to do as some other path has completed the
- * cleanup.
- */
-if (vhost->vdev) {
-data->cb(data->dev);
-}
+data->cb(data->dev);
g_free(data);
}
--
2.45.0
[PATCH v2 1/2] Revert "vhost-user: fix lost reconnect"
This reverts commit f02a4b8e6431598612466f76aac64ab492849abf.
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 2 +-
hw/scsi/vhost-user-scsi.c | 3 +--
hw/virtio/vhost-user-base.c| 2 +-
hw/virtio/vhost-user.c | 10 ++
include/hw/virtio/vhost-user.h | 3 +--
5 files changed, 6 insertions(+), 14 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 9e6bbc6950..41d1ac3a5a 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -384,7 +384,7 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent
event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >dev,
- vhost_user_blk_disconnect,
vhost_user_blk_event);
+ vhost_user_blk_disconnect);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index a63b1f4948..48a59e020e 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -214,8 +214,7 @@ static void vhost_user_scsi_event(void *opaque,
QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >conf.chardev, >dev,
- vhost_user_scsi_disconnect,
- vhost_user_scsi_event);
+ vhost_user_scsi_disconnect);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user-base.c b/hw/virtio/vhost-user-base.c
index a83167191e..4b54255682 100644
--- a/hw/virtio/vhost-user-base.c
+++ b/hw/virtio/vhost-user-base.c
@@ -254,7 +254,7 @@ static void vub_event(void *opaque, QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >vhost_dev,
- vub_disconnect, vub_event);
+ vub_disconnect);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index cdf9af4a4b..c929097e87 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2776,7 +2776,6 @@ typedef struct {
DeviceState *dev;
CharBackend *cd;
struct vhost_dev *vhost;
-IOEventHandler *event_cb;
} VhostAsyncCallback;
static void vhost_user_async_close_bh(void *opaque)
@@ -2791,10 +2790,7 @@ static void vhost_user_async_close_bh(void *opaque)
*/
if (vhost->vdev) {
data->cb(data->dev);
-} else if (data->event_cb) {
-qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
- NULL, data->dev, NULL, true);
- }
+}
g_free(data);
}
@@ -2806,8 +2802,7 @@ static void vhost_user_async_close_bh(void *opaque)
*/
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb,
-IOEventHandler *event_cb)
+vu_async_close_fn cb)
{
if (!runstate_check(RUN_STATE_SHUTDOWN)) {
/*
@@ -2823,7 +2818,6 @@ void vhost_user_async_close(DeviceState *d,
data->dev = d;
data->cd = chardev;
data->vhost = vhost;
-data->event_cb = event_cb;
/* Disable any further notifications on the chardev */
qemu_chr_fe_set_handlers(chardev,
diff --git a/include/hw/virtio/vhost-user.h b/include/hw/virtio/vhost-user.h
index d7c09ffd34..324cd8663a 100644
--- a/include/hw/virtio/vhost-user.h
+++ b/include/hw/virtio/vhost-user.h
@@ -108,7 +108,6 @@ typedef void (*vu_async_close_fn)(DeviceState *cb);
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb,
-IOEventHandler *event_cb);
+vu_async_close_fn cb);
#endif
--
2.45.0
[PATCH 1/2] Revert "vhost-user: fix lost reconnect"
This reverts commit f02a4b8e6431598612466f76aac64ab492849abf.
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 2 +-
hw/scsi/vhost-user-scsi.c | 3 +--
hw/virtio/vhost-user-base.c| 2 +-
hw/virtio/vhost-user.c | 10 ++
include/hw/virtio/vhost-user.h | 3 +--
5 files changed, 6 insertions(+), 14 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 9e6bbc6950..41d1ac3a5a 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -384,7 +384,7 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent
event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >dev,
- vhost_user_blk_disconnect,
vhost_user_blk_event);
+ vhost_user_blk_disconnect);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index a63b1f4948..48a59e020e 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -214,8 +214,7 @@ static void vhost_user_scsi_event(void *opaque,
QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >conf.chardev, >dev,
- vhost_user_scsi_disconnect,
- vhost_user_scsi_event);
+ vhost_user_scsi_disconnect);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user-base.c b/hw/virtio/vhost-user-base.c
index a83167191e..4b54255682 100644
--- a/hw/virtio/vhost-user-base.c
+++ b/hw/virtio/vhost-user-base.c
@@ -254,7 +254,7 @@ static void vub_event(void *opaque, QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >vhost_dev,
- vub_disconnect, vub_event);
+ vub_disconnect);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index cdf9af4a4b..c929097e87 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2776,7 +2776,6 @@ typedef struct {
DeviceState *dev;
CharBackend *cd;
struct vhost_dev *vhost;
-IOEventHandler *event_cb;
} VhostAsyncCallback;
static void vhost_user_async_close_bh(void *opaque)
@@ -2791,10 +2790,7 @@ static void vhost_user_async_close_bh(void *opaque)
*/
if (vhost->vdev) {
data->cb(data->dev);
-} else if (data->event_cb) {
-qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
- NULL, data->dev, NULL, true);
- }
+}
g_free(data);
}
@@ -2806,8 +2802,7 @@ static void vhost_user_async_close_bh(void *opaque)
*/
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb,
-IOEventHandler *event_cb)
+vu_async_close_fn cb)
{
if (!runstate_check(RUN_STATE_SHUTDOWN)) {
/*
@@ -2823,7 +2818,6 @@ void vhost_user_async_close(DeviceState *d,
data->dev = d;
data->cd = chardev;
data->vhost = vhost;
-data->event_cb = event_cb;
/* Disable any further notifications on the chardev */
qemu_chr_fe_set_handlers(chardev,
diff --git a/include/hw/virtio/vhost-user.h b/include/hw/virtio/vhost-user.h
index d7c09ffd34..324cd8663a 100644
--- a/include/hw/virtio/vhost-user.h
+++ b/include/hw/virtio/vhost-user.h
@@ -108,7 +108,6 @@ typedef void (*vu_async_close_fn)(DeviceState *cb);
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb,
-IOEventHandler *event_cb);
+vu_async_close_fn cb);
#endif
--
2.44.0
[PATCH 0/2] Fixed the problem of vhost-user reconnection
Previous discussion here: https://lore.kernel.org/all/f86d6159-5610-476c-a69e-cd3a717f9...@nvidia.com/ The merged version cannot fully cover all possible scenarios. Here we revert the previous fixes and then use new methods to fix them. Li Feng (2): Revert "vhost-user: fix lost reconnect" vhost-user: fix lost reconnect again hw/block/vhost-user-blk.c | 2 +- hw/scsi/vhost-user-scsi.c | 3 +-- hw/virtio/vhost-user-base.c| 2 +- hw/virtio/vhost-user.c | 18 ++ include/hw/virtio/vhost-user.h | 3 +-- 5 files changed, 6 insertions(+), 22 deletions(-) -- 2.44.0
[PATCH 2/2] vhost-user-scsi: free the inflight area when reset
Keep it the same to vhost-user-blk.
At the same time, fix the vhost_reset_device.
Signed-off-by: Li Feng
---
hw/scsi/vhost-user-scsi.c | 16
hw/virtio/virtio.c| 2 +-
2 files changed, 17 insertions(+), 1 deletion(-)
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 2060f9f94b..780f10559d 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -360,6 +360,20 @@ static Property vhost_user_scsi_properties[] = {
DEFINE_PROP_END_OF_LIST(),
};
+static void vhost_user_scsi_reset(VirtIODevice *vdev)
+{
+VHostUserSCSI *s = VHOST_USER_SCSI(vdev);
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+
+vhost_dev_free_inflight(vsc->inflight);
+}
+
+static struct vhost_dev *vhost_user_scsi_get_vhost(VirtIODevice *vdev)
+{
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(vdev);
+return >dev;
+}
+
static const VMStateDescription vmstate_vhost_scsi = {
.name = "virtio-scsi",
.minimum_version_id = 1,
@@ -385,6 +399,8 @@ static void vhost_user_scsi_class_init(ObjectClass *klass,
void *data)
vdc->set_config = vhost_scsi_common_set_config;
vdc->set_status = vhost_user_scsi_set_status;
fwc->get_dev_path = vhost_scsi_common_get_fw_dev_path;
+vdc->reset = vhost_user_scsi_reset;
+vdc->get_vhost = vhost_user_scsi_get_vhost;
}
static void vhost_user_scsi_instance_init(Object *obj)
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 4259fefeb6..d0a640af63 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -2137,7 +2137,7 @@ void virtio_reset(void *opaque)
vdev->device_endian = virtio_default_endian();
}
-if (vdev->vhost_started) {
+if (vdev->vhost_started && k->get_vhost) {
vhost_reset_device(k->get_vhost(vdev));
}
--
2.42.0
[PATCH 0/2] fix some vhost-user issues
Li Feng (2): vhost-user: fix the reconnect error vhost-user-scsi: free the inflight area when reset hw/block/vhost-user-blk.c | 8 +++- hw/scsi/vhost-user-scsi.c | 19 ++- hw/virtio/vhost-user-gpio.c | 3 ++- hw/virtio/virtio.c | 2 +- 4 files changed, 24 insertions(+), 8 deletions(-) -- 2.42.0
[PATCH 1/2] vhost-user: fix the reconnect error
If the error occurs in vhost_dev_init, the value of s->connected is set to true
in advance, and there is no chance to enter this function execution again
in the future.
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 8 +++-
hw/scsi/vhost-user-scsi.c | 3 ++-
hw/virtio/vhost-user-gpio.c | 3 ++-
3 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 818b833108..2863d80d15 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -326,7 +326,6 @@ static int vhost_user_blk_connect(DeviceState *dev, Error
**errp)
if (s->connected) {
return 0;
}
-s->connected = true;
s->dev.num_queues = s->num_queues;
s->dev.nvqs = s->num_queues;
@@ -343,15 +342,14 @@ static int vhost_user_blk_connect(DeviceState *dev, Error
**errp)
return ret;
}
+s->connected = true;
+
/* restore vhost state */
if (virtio_device_started(vdev, vdev->status)) {
ret = vhost_user_blk_start(vdev, errp);
-if (ret < 0) {
-return ret;
-}
}
-return 0;
+return ret;
}
static void vhost_user_blk_disconnect(DeviceState *dev)
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 4486500cac..2060f9f94b 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -147,7 +147,6 @@ static int vhost_user_scsi_connect(DeviceState *dev, Error
**errp)
if (s->connected) {
return 0;
}
-s->connected = true;
vsc->dev.num_queues = vs->conf.num_queues;
vsc->dev.nvqs = VIRTIO_SCSI_VQ_NUM_FIXED + vs->conf.num_queues;
@@ -161,6 +160,8 @@ static int vhost_user_scsi_connect(DeviceState *dev, Error
**errp)
return ret;
}
+s->connected = true;
+
/* restore vhost state */
if (virtio_device_started(vdev, vdev->status)) {
ret = vhost_user_scsi_start(s, errp);
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index aff2d7eff6..a83437a5da 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -229,7 +229,6 @@ static int vu_gpio_connect(DeviceState *dev, Error **errp)
if (gpio->connected) {
return 0;
}
-gpio->connected = true;
vhost_dev_set_config_notifier(vhost_dev, _ops);
gpio->vhost_user.supports_config = true;
@@ -243,6 +242,8 @@ static int vu_gpio_connect(DeviceState *dev, Error **errp)
return ret;
}
+gpio->connected = true;
+
/* restore vhost state */
if (virtio_device_started(vdev, vdev->status)) {
vu_gpio_start(vdev);
--
2.42.0
Re: [PATCH v8 0/5] Implement reconnect for vhost-user-scsi
Hello Guys, Ping… These patches have been waiting for a long time. Can they be merged? Best Regards, li > On 9 Oct 2023, at 12:46 PM, Li Feng wrote: > > Changes for v8: > - [PATCH 3/5] vhost-user-scsi: support reconnect to backend > - Fix code style suggested by Manos Pitsidianakis > - [PATCH 4/5] vhost-user-scsi: start vhost when guest kicks > - Use 'DEVICE()' macro in vhost_user_scsi_handle_output to replace the >'parent_obj.parent_obj.parent_obj.parent_obj'. > > Changes for v7: > - [PATCH 3/5] vhost-user-scsi: support reconnect to backend > - Add reporting the error in vhost-scsi; > - Rebase to master and fix the conflict. > - Add "Reviewed-by" tags. > > Changes for v6: > - [PATCH] vhost-user: fix lost reconnect > - Fix missing assign event_cb. > > Changes for v5: > - No logic has been changed, just move part of the code from patch 4 to patch > 5. > > Changes for v4: > - Merge > https://lore.kernel.org/all/20230830045722.611224-1-fen...@smartx.com/ to > this series. > - Add ERRP_GUARD in vhost_user_scsi_realize; > - Reword the commit messages. > > Changes for v3: > - Split the vhost_user_scsi_handle_output to a separate patch; > - Move the started_vu from vhost scsi common header to vhost-user-scsi header; > - Fix a log print error; > > Changes for v2: > - Split the v1 patch to small separate patchset; > - New patch for fixing fd leak, which has sent to reviewers in another > mail; > - Implement the `vhost_user_scsi_handle_output`; > - Add the started_vu safe check; > - Fix error handler; > - Check the inflight before set/get inflight fd. > > Li Feng (5): > vhost-user-common: send get_inflight_fd once > vhost: move and rename the conn retry times > vhost-user-scsi: support reconnect to backend > vhost-user-scsi: start vhost when guest kicks > vhost-user: fix lost reconnect > > hw/block/vhost-user-blk.c | 6 +- > hw/scsi/vhost-scsi-common.c | 47 ++--- > hw/scsi/vhost-scsi.c | 6 +- > hw/scsi/vhost-user-scsi.c | 250 +++--- > hw/virtio/vhost-user-gpio.c | 5 +- > hw/virtio/vhost-user.c| 10 +- > include/hw/virtio/vhost-scsi-common.h | 2 +- > include/hw/virtio/vhost-user-scsi.h | 6 + > include/hw/virtio/vhost-user.h| 3 +- > include/hw/virtio/vhost.h | 2 + > 10 files changed, 277 insertions(+), 60 deletions(-) > > -- > 2.41.0 >
[PATCH v8 3/5] vhost-user-scsi: support reconnect to backend
If the backend crashes and restarts, the device is broken.
This patch adds reconnect for vhost-user-scsi.
This patch also improves the error messages, and reports some silent errors.
Tested with spdk backend.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 16 +-
hw/scsi/vhost-scsi.c | 6 +-
hw/scsi/vhost-user-scsi.c | 201 +++---
include/hw/virtio/vhost-scsi-common.h | 2 +-
include/hw/virtio/vhost-user-scsi.h | 6 +
5 files changed, 201 insertions(+), 30 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a61cd0e907..4c8637045d 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -16,6 +16,7 @@
*/
#include "qemu/osdep.h"
+#include "qapi/error.h"
#include "qemu/error-report.h"
#include "qemu/module.h"
#include "hw/virtio/vhost.h"
@@ -25,7 +26,7 @@
#include "hw/virtio/virtio-access.h"
#include "hw/fw-path-provider.h"
-int vhost_scsi_common_start(VHostSCSICommon *vsc)
+int vhost_scsi_common_start(VHostSCSICommon *vsc, Error **errp)
{
int ret, i;
VirtIODevice *vdev = VIRTIO_DEVICE(vsc);
@@ -35,18 +36,19 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
VirtIOSCSICommon *vs = (VirtIOSCSICommon *)vsc;
if (!k->set_guest_notifiers) {
-error_report("binding does not support guest notifiers");
+error_setg(errp, "binding does not support guest notifiers");
return -ENOSYS;
}
ret = vhost_dev_enable_notifiers(>dev, vdev);
if (ret < 0) {
+error_setg_errno(errp, -ret, "Error enabling host notifiers");
return ret;
}
ret = k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, true);
if (ret < 0) {
-error_report("Error binding guest notifier");
+error_setg_errno(errp, -ret, "Error binding guest notifier");
goto err_host_notifiers;
}
@@ -54,7 +56,7 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error setting inflight format: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight format");
goto err_guest_notifiers;
}
@@ -64,21 +66,21 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vs->conf.virtqueue_size,
vsc->inflight);
if (ret < 0) {
-error_report("Error getting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error getting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_set_inflight(>dev, vsc->inflight);
if (ret < 0) {
-error_report("Error setting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_start(>dev, vdev, true);
if (ret < 0) {
-error_report("Error start vhost dev");
+error_setg_errno(errp, -ret, "Error starting vhost dev");
goto err_guest_notifiers;
}
diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c
index 443f67daa4..95cadb93e7 100644
--- a/hw/scsi/vhost-scsi.c
+++ b/hw/scsi/vhost-scsi.c
@@ -75,6 +75,7 @@ static int vhost_scsi_start(VHostSCSI *s)
int ret, abi_version;
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
const VhostOps *vhost_ops = vsc->dev.vhost_ops;
+Error *local_err = NULL;
ret = vhost_ops->vhost_scsi_get_abi_version(>dev, _version);
if (ret < 0) {
@@ -88,14 +89,15 @@ static int vhost_scsi_start(VHostSCSI *s)
return -ENOSYS;
}
-ret = vhost_scsi_common_start(vsc);
+ret = vhost_scsi_common_start(vsc, _err);
if (ret < 0) {
+error_reportf_err(local_err, "Error starting vhost-scsi");
return ret;
}
ret = vhost_scsi_set_endpoint(s);
if (ret < 0) {
-error_report("Error setting vhost-scsi endpoint");
+error_reportf_err(local_err, "Error setting vhost-scsi endpoint");
vhost_scsi_common_stop(vsc);
}
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index df6b66cc1a..1e5d853cdd 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -39,26 +39,56 @@ static const int user_feature_bits[] = {
VHOST_INVALID_FEATURE_BIT
};
+static int vhost_user_scsi_start(VHostUserSCSI *s, Error **errp)
+{
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+int ret;
+
+ret = vhost_scsi_common_start(vsc, errp);
+s->started_vu = !(ret < 0);
+
+return ret;
+
[PATCH v8 5/5] vhost-user: fix lost reconnect
When the vhost-user is reconnecting to the backend, and if the vhost-user fails
at the get_features in vhost_dev_init(), then the reconnect will fail
and it will not be retriggered forever.
The reason is:
When the vhost-user fails at get_features, the vhost_dev_cleanup will be called
immediately.
vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
The reconnect path is:
vhost_user_blk_event
vhost_user_async_close(.. vhost_user_blk_disconnect ..)
qemu_chr_fe_set_handlers <- clear the notifier callback
schedule vhost_user_async_close_bh
The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
called, then the event fd callback will not be reinstalled.
All vhost-user devices have this issue, including vhost-user-blk/scsi.
With this patch, if the vdev->vdev is null, the fd callback will still
be reinstalled.
Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/block/vhost-user-blk.c | 2 +-
hw/scsi/vhost-user-scsi.c | 3 ++-
hw/virtio/vhost-user-gpio.c| 2 +-
hw/virtio/vhost-user.c | 10 --
include/hw/virtio/vhost-user.h | 3 ++-
5 files changed, 14 insertions(+), 6 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 3c69fa47d5..95c758200d 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -391,7 +391,7 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent
event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >dev,
- vhost_user_blk_disconnect);
+ vhost_user_blk_disconnect,
vhost_user_blk_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 2457b950e1..61e4050682 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -231,7 +231,8 @@ static void vhost_user_scsi_event(void *opaque,
QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >conf.chardev, >dev,
- vhost_user_scsi_disconnect);
+ vhost_user_scsi_disconnect,
+ vhost_user_scsi_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index fc784e4213..aff2d7eff6 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -289,7 +289,7 @@ static void vu_gpio_event(void *opaque, QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >vhost_dev,
- vu_gpio_disconnect);
+ vu_gpio_disconnect, vu_gpio_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index 3766b415f8..7395bfc531 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2765,6 +2765,7 @@ typedef struct {
DeviceState *dev;
CharBackend *cd;
struct vhost_dev *vhost;
+IOEventHandler *event_cb;
} VhostAsyncCallback;
static void vhost_user_async_close_bh(void *opaque)
@@ -2779,7 +2780,10 @@ static void vhost_user_async_close_bh(void *opaque)
*/
if (vhost->vdev) {
data->cb(data->dev);
-}
+} else if (data->event_cb) {
+qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
+ NULL, data->dev, NULL, true);
+ }
g_free(data);
}
@@ -2791,7 +2795,8 @@ static void vhost_user_async_close_bh(void *opaque)
*/
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb)
+vu_async_close_fn cb,
+IOEventHandler *event_cb)
{
if (!runstate_check(RUN_STATE_SHUTDOWN)) {
/*
@@ -2807,6 +2812,7 @@ void vhost_user_async_close(DeviceState *d,
data->dev = d;
data->cd = chardev;
data->vhost = vhost;
+data->event_cb = event_cb;
/* Disable any further notifications on the chardev */
qemu_chr_fe_set_handlers(chardev,
diff --git a/include/hw/virtio/vhost-user.h b/include/hw/virtio/vhost-user.h
index 9f9ddf878d..6b06ecb1bd 100644
--- a/include/hw/virtio/vhost-user.h
+++ b/include/hw/virtio/vhost-user.h
@@ -106,6 +106,7 @@ typedef void (*vu_async_close_fn)(DeviceState *cb);
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, stru
[PATCH v8 2/5] vhost: move and rename the conn retry times
Multiple devices need this macro, move it to a common header.
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/block/vhost-user-blk.c | 4 +---
hw/virtio/vhost-user-gpio.c | 3 +--
include/hw/virtio/vhost.h | 2 ++
3 files changed, 4 insertions(+), 5 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index eecf3f7a81..3c69fa47d5 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -32,8 +32,6 @@
#include "sysemu/sysemu.h"
#include "sysemu/runstate.h"
-#define REALIZE_CONNECTION_RETRIES 3
-
static const int user_feature_bits[] = {
VIRTIO_BLK_F_SIZE_MAX,
VIRTIO_BLK_F_SEG_MAX,
@@ -482,7 +480,7 @@ static void vhost_user_blk_device_realize(DeviceState *dev,
Error **errp)
s->inflight = g_new0(struct vhost_inflight, 1);
s->vhost_vqs = g_new0(struct vhost_virtqueue, s->num_queues);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
assert(!*errp);
do {
if (*errp) {
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index 3d7fae3984..fc784e4213 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -15,7 +15,6 @@
#include "standard-headers/linux/virtio_ids.h"
#include "trace.h"
-#define REALIZE_CONNECTION_RETRIES 3
#define VHOST_NVQS 2
/* Features required from VirtIO */
@@ -365,7 +364,7 @@ static void vu_gpio_device_realize(DeviceState *dev, Error
**errp)
qemu_chr_fe_set_handlers(>chardev, NULL, NULL, vu_gpio_event, NULL,
dev, NULL, true);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
g_assert(!*errp);
do {
if (*errp) {
diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h
index 6a173cb9fa..ca3131b1af 100644
--- a/include/hw/virtio/vhost.h
+++ b/include/hw/virtio/vhost.h
@@ -8,6 +8,8 @@
#define VHOST_F_DEVICE_IOTLB 63
#define VHOST_USER_F_PROTOCOL_FEATURES 30
+#define VU_REALIZE_CONN_RETRIES 3
+
/* Generic structures common for any vhost based device. */
struct vhost_inflight {
--
2.41.0
[PATCH v8 4/5] vhost-user-scsi: start vhost when guest kicks
Let's keep the same behavior as vhost-user-blk.
Some old guests kick virtqueue before setting VIRTIO_CONFIG_S_DRIVER_OK.
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/scsi/vhost-user-scsi.c | 48 +++
1 file changed, 44 insertions(+), 4 deletions(-)
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 1e5d853cdd..2457b950e1 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -111,8 +111,48 @@ static void vhost_user_scsi_reset(VirtIODevice *vdev)
}
}
-static void vhost_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+static void vhost_user_scsi_handle_output(VirtIODevice *vdev, VirtQueue *vq)
{
+VHostUserSCSI *s = (VHostUserSCSI *)vdev;
+DeviceState *dev = DEVICE(vdev);
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+
+Error *local_err = NULL;
+int i, ret;
+
+if (!vdev->start_on_kick) {
+return;
+}
+
+if (!s->connected) {
+return;
+}
+
+if (vhost_dev_is_started(>dev)) {
+return;
+}
+
+/*
+ * Some guests kick before setting VIRTIO_CONFIG_S_DRIVER_OK so start
+ * vhost here instead of waiting for .set_status().
+ */
+ret = vhost_user_scsi_start(s, _err);
+if (ret < 0) {
+error_reportf_err(local_err, "vhost-user-scsi: vhost start failed: ");
+qemu_chr_fe_disconnect(>conf.chardev);
+return;
+}
+
+/* Kick right away to begin processing requests already in vring */
+for (i = 0; i < vsc->dev.nvqs; i++) {
+VirtQueue *kick_vq = virtio_get_queue(vdev, i);
+
+if (!virtio_queue_get_desc_addr(vdev, i)) {
+continue;
+}
+event_notifier_set(virtio_queue_get_host_notifier(kick_vq));
+}
}
static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
@@ -239,9 +279,9 @@ static void vhost_user_scsi_realize(DeviceState *dev, Error
**errp)
return;
}
-virtio_scsi_common_realize(dev, vhost_dummy_handle_output,
- vhost_dummy_handle_output,
- vhost_dummy_handle_output, );
+virtio_scsi_common_realize(dev, vhost_user_scsi_handle_output,
+ vhost_user_scsi_handle_output,
+ vhost_user_scsi_handle_output, );
if (err != NULL) {
error_propagate(errp, err);
return;
--
2.41.0
[PATCH v8 1/5] vhost-user-common: send get_inflight_fd once
Currently the get_inflight_fd will be sent every time the device is started, and
the backend will allocate shared memory to save the inflight state. If the
backend finds that it receives the second get_inflight_fd, it will release the
previous shared memory, which breaks inflight working logic.
This patch is a preparation for the following patches.
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/scsi/vhost-scsi-common.c | 37 ++---
1 file changed, 18 insertions(+), 19 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a06f01af26..a61cd0e907 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -52,20 +52,28 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vsc->dev.acked_features = vdev->guest_features;
-assert(vsc->inflight == NULL);
-vsc->inflight = g_new0(struct vhost_inflight, 1);
-ret = vhost_dev_get_inflight(>dev,
- vs->conf.virtqueue_size,
- vsc->inflight);
+ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error get inflight: %d", -ret);
+error_report("Error setting inflight format: %d", -ret);
goto err_guest_notifiers;
}
-ret = vhost_dev_set_inflight(>dev, vsc->inflight);
-if (ret < 0) {
-error_report("Error set inflight: %d", -ret);
-goto err_guest_notifiers;
+if (vsc->inflight) {
+if (!vsc->inflight->addr) {
+ret = vhost_dev_get_inflight(>dev,
+vs->conf.virtqueue_size,
+vsc->inflight);
+if (ret < 0) {
+error_report("Error getting inflight: %d", -ret);
+goto err_guest_notifiers;
+}
+}
+
+ret = vhost_dev_set_inflight(>dev, vsc->inflight);
+if (ret < 0) {
+error_report("Error setting inflight: %d", -ret);
+goto err_guest_notifiers;
+}
}
ret = vhost_dev_start(>dev, vdev, true);
@@ -85,9 +93,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
return ret;
err_guest_notifiers:
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-
k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
err_host_notifiers:
vhost_dev_disable_notifiers(>dev, vdev);
@@ -111,12 +116,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
}
assert(ret >= 0);
-if (vsc->inflight) {
-vhost_dev_free_inflight(vsc->inflight);
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-}
-
vhost_dev_disable_notifiers(>dev, vdev);
}
--
2.41.0
[PATCH v8 0/5] Implement reconnect for vhost-user-scsi
Changes for v8: - [PATCH 3/5] vhost-user-scsi: support reconnect to backend - Fix code style suggested by Manos Pitsidianakis - [PATCH 4/5] vhost-user-scsi: start vhost when guest kicks - Use 'DEVICE()' macro in vhost_user_scsi_handle_output to replace the 'parent_obj.parent_obj.parent_obj.parent_obj'. Changes for v7: - [PATCH 3/5] vhost-user-scsi: support reconnect to backend - Add reporting the error in vhost-scsi; - Rebase to master and fix the conflict. - Add "Reviewed-by" tags. Changes for v6: - [PATCH] vhost-user: fix lost reconnect - Fix missing assign event_cb. Changes for v5: - No logic has been changed, just move part of the code from patch 4 to patch 5. Changes for v4: - Merge https://lore.kernel.org/all/20230830045722.611224-1-fen...@smartx.com/ to this series. - Add ERRP_GUARD in vhost_user_scsi_realize; - Reword the commit messages. Changes for v3: - Split the vhost_user_scsi_handle_output to a separate patch; - Move the started_vu from vhost scsi common header to vhost-user-scsi header; - Fix a log print error; Changes for v2: - Split the v1 patch to small separate patchset; - New patch for fixing fd leak, which has sent to reviewers in another mail; - Implement the `vhost_user_scsi_handle_output`; - Add the started_vu safe check; - Fix error handler; - Check the inflight before set/get inflight fd. Li Feng (5): vhost-user-common: send get_inflight_fd once vhost: move and rename the conn retry times vhost-user-scsi: support reconnect to backend vhost-user-scsi: start vhost when guest kicks vhost-user: fix lost reconnect hw/block/vhost-user-blk.c | 6 +- hw/scsi/vhost-scsi-common.c | 47 ++--- hw/scsi/vhost-scsi.c | 6 +- hw/scsi/vhost-user-scsi.c | 250 +++--- hw/virtio/vhost-user-gpio.c | 5 +- hw/virtio/vhost-user.c| 10 +- include/hw/virtio/vhost-scsi-common.h | 2 +- include/hw/virtio/vhost-user-scsi.h | 6 + include/hw/virtio/vhost-user.h| 3 +- include/hw/virtio/vhost.h | 2 + 10 files changed, 277 insertions(+), 60 deletions(-) -- 2.41.0
Re: [PATCH v7 3/5] vhost-user-scsi: support reconnect to backend
Thanks for your comments, I will submit the v8.
> On 8 Oct 2023, at 6:46 PM, Manos Pitsidianakis
> wrote:
>
> Hello Li, I have some trivial style comments you could possibly address in a
> next version:
>
> On Sun, 08 Oct 2023 12:12, Li Feng wrote:
>> diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
>> index df6b66cc1a..5df24faff4 100644
>> --- a/hw/scsi/vhost-user-scsi.c
>> +++ b/hw/scsi/vhost-user-scsi.c
>> @@ -39,26 +39,56 @@ static const int user_feature_bits[] = {
>>VHOST_INVALID_FEATURE_BIT
>> };
>> +static int vhost_user_scsi_start(VHostUserSCSI *s, Error **errp)
>> +{
>> +VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
>> +int ret;
>> +
>> +ret = vhost_scsi_common_start(vsc, errp);
>> +s->started_vu = (ret < 0 ? false : true);
>
> -+s->started_vu = (ret < 0 ? false : true);
> ++s->started_vu = !(ret < 0);
>
>> static void vhost_user_scsi_set_status(VirtIODevice *vdev, uint8_t status)
>> {
>>VHostUserSCSI *s = (VHostUserSCSI *)vdev;
>> +DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
>
> -+DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
> ++DeviceState *dev = DEVICE(vdev);
>
>> +static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
>> +{
>> +VirtIODevice *vdev = VIRTIO_DEVICE(dev);
>> +VHostUserSCSI *s = VHOST_USER_SCSI(vdev);
>> +VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
>> +VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
>> +int ret = 0;
>> +
>> +if (s->connected) {
>> +return 0;
>> +}
>> +s->connected = true;
>> +
>> +vsc->dev.num_queues = vs->conf.num_queues;
>> +vsc->dev.nvqs = VIRTIO_SCSI_VQ_NUM_FIXED + vs->conf.num_queues;
>> +vsc->dev.vqs = s->vhost_vqs;
>> +vsc->dev.vq_index = 0;
>> +vsc->dev.backend_features = 0;
>> +
>> +ret = vhost_dev_init(>dev, >vhost_user,
>> VHOST_BACKEND_TYPE_USER, 0,
>> + errp);
>> +if (ret < 0) {
>> +return ret;
>> +}
>> +
>> +/* restore vhost state */
>> +if (virtio_device_started(vdev, vdev->status)) {
>> +ret = vhost_user_scsi_start(s, errp);
>> +if (ret < 0) {
>> +return ret;
>> +}
>> +}
>> +
>> +return 0;
>> +}
>
>
> -+if (virtio_device_started(vdev, vdev->status)) {
> -+ret = vhost_user_scsi_start(s, errp);
> -+if (ret < 0) {
> -+return ret;
> -+}
> -+}
> -+
> -+return 0;
> -+}
> ++if (virtio_device_started(vdev, vdev->status)) {
> ++ret = vhost_user_scsi_start(s, errp);
> ++}
> ++
> ++return ret;
> ++}
>
> [skipping..]
>
>> +static int vhost_user_scsi_realize_connect(VHostUserSCSI *s, Error **errp)
>> +{
>> +DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
>
>
> -+DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
> ++DeviceState *dev = DEVICE(s);
>
>> diff --git a/include/hw/virtio/vhost-user-scsi.h
>> b/include/hw/virtio/vhost-user-scsi.h
>> index 521b08e559..b405ec952a 100644
>> --- a/include/hw/virtio/vhost-user-scsi.h
>> +++ b/include/hw/virtio/vhost-user-scsi.h
>> @@ -29,6 +29,10 @@ OBJECT_DECLARE_SIMPLE_TYPE(VHostUserSCSI, VHOST_USER_SCSI)
>> struct VHostUserSCSI {
>>VHostSCSICommon parent_obj;
>>VhostUserState vhost_user;
>> +bool connected;
>> +bool started_vu;
>> +
>> +struct vhost_virtqueue *vhost_vqs;
>
> +bool connected;
> +bool started_vu;
> -+
> +struct vhost_virtqueue *vhost_vqs;
>
> See
> https://www.qemu.org/docs/master/devel/style.html#qemu-object-model-declarations
>
> The definition should look like:
>
> struct VHostUserSCSI {
> VHostSCSICommon parent_obj;
>
> /* Properties */
> bool connected;
> bool started_vu;
>
> VhostUserState vhost_user;
> struct vhost_virtqueue *vhost_vqs;
> }
[PATCH v7 1/5] vhost-user-common: send get_inflight_fd once
Currently the get_inflight_fd will be sent every time the device is started, and
the backend will allocate shared memory to save the inflight state. If the
backend finds that it receives the second get_inflight_fd, it will release the
previous shared memory, which breaks inflight working logic.
This patch is a preparation for the following patches.
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/scsi/vhost-scsi-common.c | 37 ++---
1 file changed, 18 insertions(+), 19 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a06f01af26..a61cd0e907 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -52,20 +52,28 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vsc->dev.acked_features = vdev->guest_features;
-assert(vsc->inflight == NULL);
-vsc->inflight = g_new0(struct vhost_inflight, 1);
-ret = vhost_dev_get_inflight(>dev,
- vs->conf.virtqueue_size,
- vsc->inflight);
+ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error get inflight: %d", -ret);
+error_report("Error setting inflight format: %d", -ret);
goto err_guest_notifiers;
}
-ret = vhost_dev_set_inflight(>dev, vsc->inflight);
-if (ret < 0) {
-error_report("Error set inflight: %d", -ret);
-goto err_guest_notifiers;
+if (vsc->inflight) {
+if (!vsc->inflight->addr) {
+ret = vhost_dev_get_inflight(>dev,
+vs->conf.virtqueue_size,
+vsc->inflight);
+if (ret < 0) {
+error_report("Error getting inflight: %d", -ret);
+goto err_guest_notifiers;
+}
+}
+
+ret = vhost_dev_set_inflight(>dev, vsc->inflight);
+if (ret < 0) {
+error_report("Error setting inflight: %d", -ret);
+goto err_guest_notifiers;
+}
}
ret = vhost_dev_start(>dev, vdev, true);
@@ -85,9 +93,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
return ret;
err_guest_notifiers:
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-
k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
err_host_notifiers:
vhost_dev_disable_notifiers(>dev, vdev);
@@ -111,12 +116,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
}
assert(ret >= 0);
-if (vsc->inflight) {
-vhost_dev_free_inflight(vsc->inflight);
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-}
-
vhost_dev_disable_notifiers(>dev, vdev);
}
--
2.41.0
[PATCH v7 5/5] vhost-user: fix lost reconnect
When the vhost-user is reconnecting to the backend, and if the vhost-user fails
at the get_features in vhost_dev_init(), then the reconnect will fail
and it will not be retriggered forever.
The reason is:
When the vhost-user fails at get_features, the vhost_dev_cleanup will be called
immediately.
vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
The reconnect path is:
vhost_user_blk_event
vhost_user_async_close(.. vhost_user_blk_disconnect ..)
qemu_chr_fe_set_handlers <- clear the notifier callback
schedule vhost_user_async_close_bh
The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
called, then the event fd callback will not be reinstalled.
All vhost-user devices have this issue, including vhost-user-blk/scsi.
With this patch, if the vdev->vdev is null, the fd callback will still
be reinstalled.
Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/block/vhost-user-blk.c | 2 +-
hw/scsi/vhost-user-scsi.c | 3 ++-
hw/virtio/vhost-user-gpio.c| 2 +-
hw/virtio/vhost-user.c | 10 --
include/hw/virtio/vhost-user.h | 3 ++-
5 files changed, 14 insertions(+), 6 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 3c69fa47d5..95c758200d 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -391,7 +391,7 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent
event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >dev,
- vhost_user_blk_disconnect);
+ vhost_user_blk_disconnect,
vhost_user_blk_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 5afb514398..dbe864c0e5 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -234,7 +234,8 @@ static void vhost_user_scsi_event(void *opaque,
QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >conf.chardev, >dev,
- vhost_user_scsi_disconnect);
+ vhost_user_scsi_disconnect,
+ vhost_user_scsi_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index fc784e4213..aff2d7eff6 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -289,7 +289,7 @@ static void vu_gpio_event(void *opaque, QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >vhost_dev,
- vu_gpio_disconnect);
+ vu_gpio_disconnect, vu_gpio_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index 3766b415f8..7395bfc531 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2765,6 +2765,7 @@ typedef struct {
DeviceState *dev;
CharBackend *cd;
struct vhost_dev *vhost;
+IOEventHandler *event_cb;
} VhostAsyncCallback;
static void vhost_user_async_close_bh(void *opaque)
@@ -2779,7 +2780,10 @@ static void vhost_user_async_close_bh(void *opaque)
*/
if (vhost->vdev) {
data->cb(data->dev);
-}
+} else if (data->event_cb) {
+qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
+ NULL, data->dev, NULL, true);
+ }
g_free(data);
}
@@ -2791,7 +2795,8 @@ static void vhost_user_async_close_bh(void *opaque)
*/
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb)
+vu_async_close_fn cb,
+IOEventHandler *event_cb)
{
if (!runstate_check(RUN_STATE_SHUTDOWN)) {
/*
@@ -2807,6 +2812,7 @@ void vhost_user_async_close(DeviceState *d,
data->dev = d;
data->cd = chardev;
data->vhost = vhost;
+data->event_cb = event_cb;
/* Disable any further notifications on the chardev */
qemu_chr_fe_set_handlers(chardev,
diff --git a/include/hw/virtio/vhost-user.h b/include/hw/virtio/vhost-user.h
index 9f9ddf878d..6b06ecb1bd 100644
--- a/include/hw/virtio/vhost-user.h
+++ b/include/hw/virtio/vhost-user.h
@@ -106,6 +106,7 @@ typedef void (*vu_async_close_fn)(DeviceState *cb);
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, stru
[PATCH v7 4/5] vhost-user-scsi: start vhost when guest kicks
Let's keep the same behavior as vhost-user-blk.
Some old guests kick virtqueue before setting VIRTIO_CONFIG_S_DRIVER_OK.
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/scsi/vhost-user-scsi.c | 48 +++
1 file changed, 44 insertions(+), 4 deletions(-)
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 5df24faff4..5afb514398 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -111,8 +111,48 @@ static void vhost_user_scsi_reset(VirtIODevice *vdev)
}
}
-static void vhost_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+static void vhost_user_scsi_handle_output(VirtIODevice *vdev, VirtQueue *vq)
{
+VHostUserSCSI *s = (VHostUserSCSI *)vdev;
+DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+
+Error *local_err = NULL;
+int i, ret;
+
+if (!vdev->start_on_kick) {
+return;
+}
+
+if (!s->connected) {
+return;
+}
+
+if (vhost_dev_is_started(>dev)) {
+return;
+}
+
+/*
+ * Some guests kick before setting VIRTIO_CONFIG_S_DRIVER_OK so start
+ * vhost here instead of waiting for .set_status().
+ */
+ret = vhost_user_scsi_start(s, _err);
+if (ret < 0) {
+error_reportf_err(local_err, "vhost-user-scsi: vhost start failed: ");
+qemu_chr_fe_disconnect(>conf.chardev);
+return;
+}
+
+/* Kick right away to begin processing requests already in vring */
+for (i = 0; i < vsc->dev.nvqs; i++) {
+VirtQueue *kick_vq = virtio_get_queue(vdev, i);
+
+if (!virtio_queue_get_desc_addr(vdev, i)) {
+continue;
+}
+event_notifier_set(virtio_queue_get_host_notifier(kick_vq));
+}
}
static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
@@ -242,9 +282,9 @@ static void vhost_user_scsi_realize(DeviceState *dev, Error
**errp)
return;
}
-virtio_scsi_common_realize(dev, vhost_dummy_handle_output,
- vhost_dummy_handle_output,
- vhost_dummy_handle_output, );
+virtio_scsi_common_realize(dev, vhost_user_scsi_handle_output,
+ vhost_user_scsi_handle_output,
+ vhost_user_scsi_handle_output, );
if (err != NULL) {
error_propagate(errp, err);
return;
--
2.41.0
[PATCH v7 3/5] vhost-user-scsi: support reconnect to backend
If the backend crashes and restarts, the device is broken.
This patch adds reconnect for vhost-user-scsi.
This patch also improves the error messages, and reports some silent errors.
Tested with spdk backend.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 16 +-
hw/scsi/vhost-scsi.c | 6 +-
hw/scsi/vhost-user-scsi.c | 204 +++---
include/hw/virtio/vhost-scsi-common.h | 2 +-
include/hw/virtio/vhost-user-scsi.h | 4 +
5 files changed, 202 insertions(+), 30 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a61cd0e907..4c8637045d 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -16,6 +16,7 @@
*/
#include "qemu/osdep.h"
+#include "qapi/error.h"
#include "qemu/error-report.h"
#include "qemu/module.h"
#include "hw/virtio/vhost.h"
@@ -25,7 +26,7 @@
#include "hw/virtio/virtio-access.h"
#include "hw/fw-path-provider.h"
-int vhost_scsi_common_start(VHostSCSICommon *vsc)
+int vhost_scsi_common_start(VHostSCSICommon *vsc, Error **errp)
{
int ret, i;
VirtIODevice *vdev = VIRTIO_DEVICE(vsc);
@@ -35,18 +36,19 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
VirtIOSCSICommon *vs = (VirtIOSCSICommon *)vsc;
if (!k->set_guest_notifiers) {
-error_report("binding does not support guest notifiers");
+error_setg(errp, "binding does not support guest notifiers");
return -ENOSYS;
}
ret = vhost_dev_enable_notifiers(>dev, vdev);
if (ret < 0) {
+error_setg_errno(errp, -ret, "Error enabling host notifiers");
return ret;
}
ret = k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, true);
if (ret < 0) {
-error_report("Error binding guest notifier");
+error_setg_errno(errp, -ret, "Error binding guest notifier");
goto err_host_notifiers;
}
@@ -54,7 +56,7 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error setting inflight format: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight format");
goto err_guest_notifiers;
}
@@ -64,21 +66,21 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vs->conf.virtqueue_size,
vsc->inflight);
if (ret < 0) {
-error_report("Error getting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error getting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_set_inflight(>dev, vsc->inflight);
if (ret < 0) {
-error_report("Error setting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_start(>dev, vdev, true);
if (ret < 0) {
-error_report("Error start vhost dev");
+error_setg_errno(errp, -ret, "Error starting vhost dev");
goto err_guest_notifiers;
}
diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c
index 443f67daa4..95cadb93e7 100644
--- a/hw/scsi/vhost-scsi.c
+++ b/hw/scsi/vhost-scsi.c
@@ -75,6 +75,7 @@ static int vhost_scsi_start(VHostSCSI *s)
int ret, abi_version;
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
const VhostOps *vhost_ops = vsc->dev.vhost_ops;
+Error *local_err = NULL;
ret = vhost_ops->vhost_scsi_get_abi_version(>dev, _version);
if (ret < 0) {
@@ -88,14 +89,15 @@ static int vhost_scsi_start(VHostSCSI *s)
return -ENOSYS;
}
-ret = vhost_scsi_common_start(vsc);
+ret = vhost_scsi_common_start(vsc, _err);
if (ret < 0) {
+error_reportf_err(local_err, "Error starting vhost-scsi");
return ret;
}
ret = vhost_scsi_set_endpoint(s);
if (ret < 0) {
-error_report("Error setting vhost-scsi endpoint");
+error_reportf_err(local_err, "Error setting vhost-scsi endpoint");
vhost_scsi_common_stop(vsc);
}
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index df6b66cc1a..5df24faff4 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -39,26 +39,56 @@ static const int user_feature_bits[] = {
VHOST_INVALID_FEATURE_BIT
};
+static int vhost_user_scsi_start(VHostUserSCSI *s, Error **errp)
+{
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+int ret;
+
+ret = vhost_scsi_common_start(vsc, errp);
+s->started_vu = (ret < 0 ? false : true);
+
+
[PATCH v7 2/5] vhost: move and rename the conn retry times
Multiple devices need this macro, move it to a common header.
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/block/vhost-user-blk.c | 4 +---
hw/virtio/vhost-user-gpio.c | 3 +--
include/hw/virtio/vhost.h | 2 ++
3 files changed, 4 insertions(+), 5 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index eecf3f7a81..3c69fa47d5 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -32,8 +32,6 @@
#include "sysemu/sysemu.h"
#include "sysemu/runstate.h"
-#define REALIZE_CONNECTION_RETRIES 3
-
static const int user_feature_bits[] = {
VIRTIO_BLK_F_SIZE_MAX,
VIRTIO_BLK_F_SEG_MAX,
@@ -482,7 +480,7 @@ static void vhost_user_blk_device_realize(DeviceState *dev,
Error **errp)
s->inflight = g_new0(struct vhost_inflight, 1);
s->vhost_vqs = g_new0(struct vhost_virtqueue, s->num_queues);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
assert(!*errp);
do {
if (*errp) {
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index 3d7fae3984..fc784e4213 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -15,7 +15,6 @@
#include "standard-headers/linux/virtio_ids.h"
#include "trace.h"
-#define REALIZE_CONNECTION_RETRIES 3
#define VHOST_NVQS 2
/* Features required from VirtIO */
@@ -365,7 +364,7 @@ static void vu_gpio_device_realize(DeviceState *dev, Error
**errp)
qemu_chr_fe_set_handlers(>chardev, NULL, NULL, vu_gpio_event, NULL,
dev, NULL, true);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
g_assert(!*errp);
do {
if (*errp) {
diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h
index 6a173cb9fa..ca3131b1af 100644
--- a/include/hw/virtio/vhost.h
+++ b/include/hw/virtio/vhost.h
@@ -8,6 +8,8 @@
#define VHOST_F_DEVICE_IOTLB 63
#define VHOST_USER_F_PROTOCOL_FEATURES 30
+#define VU_REALIZE_CONN_RETRIES 3
+
/* Generic structures common for any vhost based device. */
struct vhost_inflight {
--
2.41.0
[PATCH v7 0/5] Implement reconnect for vhost-user-scsi
Changes for v7: - [PATCH 3/5] vhost-user-scsi: support reconnect to backend - Add reporting the error in vhost-scsi; - Rebase to master and fix the conflict. - Add "Reviewed-by" tags. Changes for v6: - [PATCH] vhost-user: fix lost reconnect - Fix missing assign event_cb. Changes for v5: - No logic has been changed, just move part of the code from patch 4 to patch 5. Changes for v4: - Merge https://lore.kernel.org/all/20230830045722.611224-1-fen...@smartx.com/ to this series. - Add ERRP_GUARD in vhost_user_scsi_realize; - Reword the commit messages. Changes for v3: - Split the vhost_user_scsi_handle_output to a separate patch; - Move the started_vu from vhost scsi common header to vhost-user-scsi header; - Fix a log print error; Changes for v2: - Split the v1 patch to small separate patchset; - New patch for fixing fd leak, which has sent to reviewers in another mail; - Implement the `vhost_user_scsi_handle_output`; - Add the started_vu safe check; - Fix error handler; - Check the inflight before set/get inflight fd. Li Feng (5): vhost-user-common: send get_inflight_fd once vhost: move and rename the conn retry times vhost-user-scsi: support reconnect to backend vhost-user-scsi: start vhost when guest kicks vhost-user: fix lost reconnect hw/block/vhost-user-blk.c | 6 +- hw/scsi/vhost-scsi-common.c | 47 ++--- hw/scsi/vhost-scsi.c | 6 +- hw/scsi/vhost-user-scsi.c | 253 +++--- hw/virtio/vhost-user-gpio.c | 5 +- hw/virtio/vhost-user.c| 10 +- include/hw/virtio/vhost-scsi-common.h | 2 +- include/hw/virtio/vhost-user-scsi.h | 4 + include/hw/virtio/vhost-user.h| 3 +- include/hw/virtio/vhost.h | 2 + 10 files changed, 278 insertions(+), 60 deletions(-) -- 2.41.0
Re: [PATCH v6 0/5] Implement reconnect for vhost-user-scsi
On Sun, Oct 8, 2023 at 4:49 PM Michael S. Tsirkin wrote: > > On Fri, Sep 22, 2023 at 07:46:10PM +0800, Li Feng wrote: > > Changes for v6: > > - [PATCH] vhost-user: fix lost reconnect > > - Fix missing assign event_cb. > > > Pls don't make vN+1 a reply to vN - start a new thread > with each version please. OK. > > > Changes for v5: > > - No logic has been changed, just move part of the code from patch 4 to > > patch 5. > > > > Changes for v4: > > - Merge > > https://lore.kernel.org/all/20230830045722.611224-1-fen...@smartx.com/ to > > this series. > > - Add ERRP_GUARD in vhost_user_scsi_realize; > > - Reword the commit messages. > > > > Changes for v3: > > - Split the vhost_user_scsi_handle_output to a separate patch; > > - Move the started_vu from vhost scsi common header to vhost-user-scsi > > header; > > - Fix a log print error; > > > > Changes for v2: > > - Split the v1 patch to small separate patchset; > > - New patch for fixing fd leak, which has sent to reviewers in another > > mail; > > - Implement the `vhost_user_scsi_handle_output`; > > - Add the started_vu safe check; > > - Fix error handler; > > - Check the inflight before set/get inflight fd. > > > > Li Feng (5): > > vhost-user-common: send get_inflight_fd once > > vhost: move and rename the conn retry times > > vhost-user-scsi: support reconnect to backend > > vhost-user-scsi: start vhost when guest kicks > > vhost-user: fix lost reconnect > > > > hw/block/vhost-user-blk.c | 6 +- > > hw/scsi/vhost-scsi-common.c | 47 ++--- > > hw/scsi/vhost-scsi.c | 5 +- > > hw/scsi/vhost-user-scsi.c | 253 +++--- > > hw/virtio/vhost-user-gpio.c | 5 +- > > hw/virtio/vhost-user.c| 10 +- > > include/hw/virtio/vhost-scsi-common.h | 2 +- > > include/hw/virtio/vhost-user-scsi.h | 4 + > > include/hw/virtio/vhost-user.h| 3 +- > > include/hw/virtio/vhost.h | 2 + > > 10 files changed, 277 insertions(+), 60 deletions(-) > > > > -- > > 2.41.0 >
Re: [PATCH v6 1/5] vhost-user-common: send get_inflight_fd once
On Sun, Oct 8, 2023 at 4:51 PM Michael S. Tsirkin wrote:
>
> On Sun, Oct 08, 2023 at 04:49:05PM +0800, Li Feng wrote:
> > On Fri, Sep 29, 2023 at 8:55 AM Raphael Norwitz
> > wrote:
> > >
> > >
> > >
> > > > On Sep 22, 2023, at 7:46 AM, Li Feng wrote:
> > > >
> > > > Currently the get_inflight_fd will be sent every time the device is
> > > > started, and
> > > > the backend will allocate shared memory to save the inflight state. If
> > > > the
> > > > backend finds that it receives the second get_inflight_fd, it will
> > > > release the
> > > > previous shared memory, which breaks inflight working logic.
> > > >
> > > > This patch is a preparation for the following patches.
> > >
> > > This looks identical to the v3 patch I reviewed? If I’ve missed something
> > > can you please point it out?
> > Yes, nothing changed in this patch.
>
>
> Then you should include tags such as reviewed/acked by for previous
> version. if you drop tags you indicate to people they have to
> re-review.
> also, mentioning which patches changed in the cover letter is
> a courtesy to reviewers.
OK.
>
> > >
> > >
> > > > Signed-off-by: Li Feng
> > > > ---
> > > > hw/scsi/vhost-scsi-common.c | 37 ++---
> > > > 1 file changed, 18 insertions(+), 19 deletions(-)
> > > >
> > > > diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
> > > > index a06f01af26..a61cd0e907 100644
> > > > --- a/hw/scsi/vhost-scsi-common.c
> > > > +++ b/hw/scsi/vhost-scsi-common.c
> > > > @@ -52,20 +52,28 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
> > > >
> > > > vsc->dev.acked_features = vdev->guest_features;
> > > >
> > > > -assert(vsc->inflight == NULL);
> > > > -vsc->inflight = g_new0(struct vhost_inflight, 1);
> > > > -ret = vhost_dev_get_inflight(>dev,
> > > > - vs->conf.virtqueue_size,
> > > > - vsc->inflight);
> > > > +ret = vhost_dev_prepare_inflight(>dev, vdev);
> > > > if (ret < 0) {
> > > > -error_report("Error get inflight: %d", -ret);
> > > > +error_report("Error setting inflight format: %d", -ret);
> > > > goto err_guest_notifiers;
> > > > }
> > > >
> > > > -ret = vhost_dev_set_inflight(>dev, vsc->inflight);
> > > > -if (ret < 0) {
> > > > -error_report("Error set inflight: %d", -ret);
> > > > -goto err_guest_notifiers;
> > > > +if (vsc->inflight) {
> > > > +if (!vsc->inflight->addr) {
> > > > +ret = vhost_dev_get_inflight(>dev,
> > > > +vs->conf.virtqueue_size,
> > > > +vsc->inflight);
> > > > +if (ret < 0) {
> > > > +error_report("Error getting inflight: %d", -ret);
> > > > +goto err_guest_notifiers;
> > > > +}
> > > > +}
> > > > +
> > > > +ret = vhost_dev_set_inflight(>dev, vsc->inflight);
> > > > +if (ret < 0) {
> > > > +error_report("Error setting inflight: %d", -ret);
> > > > +goto err_guest_notifiers;
> > > > +}
> > > > }
> > > >
> > > > ret = vhost_dev_start(>dev, vdev, true);
> > > > @@ -85,9 +93,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
> > > > return ret;
> > > >
> > > > err_guest_notifiers:
> > > > -g_free(vsc->inflight);
> > > > -vsc->inflight = NULL;
> > > > -
> > > > k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
> > > > err_host_notifiers:
> > > > vhost_dev_disable_notifiers(>dev, vdev);
> > > > @@ -111,12 +116,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
> > > > }
> > > > assert(ret >= 0);
> > > >
> > > > -if (vsc->inflight) {
> > > > -vhost_dev_free_inflight(vsc->inflight);
> > > > -g_free(vsc->inflight);
> > > > -vsc->inflight = NULL;
> > > > -}
> > > > -
> > > > vhost_dev_disable_notifiers(>dev, vdev);
> > > > }
> > > >
> > > > --
> > > > 2.41.0
> > > >
> > >
>
Re: [PATCH v6 1/5] vhost-user-common: send get_inflight_fd once
On Fri, Sep 29, 2023 at 8:55 AM Raphael Norwitz
wrote:
>
>
>
> > On Sep 22, 2023, at 7:46 AM, Li Feng wrote:
> >
> > Currently the get_inflight_fd will be sent every time the device is
> > started, and
> > the backend will allocate shared memory to save the inflight state. If the
> > backend finds that it receives the second get_inflight_fd, it will release
> > the
> > previous shared memory, which breaks inflight working logic.
> >
> > This patch is a preparation for the following patches.
>
> This looks identical to the v3 patch I reviewed? If I’ve missed something can
> you please point it out?
Yes, nothing changed in this patch.
>
>
> > Signed-off-by: Li Feng
> > ---
> > hw/scsi/vhost-scsi-common.c | 37 ++---
> > 1 file changed, 18 insertions(+), 19 deletions(-)
> >
> > diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
> > index a06f01af26..a61cd0e907 100644
> > --- a/hw/scsi/vhost-scsi-common.c
> > +++ b/hw/scsi/vhost-scsi-common.c
> > @@ -52,20 +52,28 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
> >
> > vsc->dev.acked_features = vdev->guest_features;
> >
> > -assert(vsc->inflight == NULL);
> > -vsc->inflight = g_new0(struct vhost_inflight, 1);
> > -ret = vhost_dev_get_inflight(>dev,
> > - vs->conf.virtqueue_size,
> > - vsc->inflight);
> > +ret = vhost_dev_prepare_inflight(>dev, vdev);
> > if (ret < 0) {
> > -error_report("Error get inflight: %d", -ret);
> > +error_report("Error setting inflight format: %d", -ret);
> > goto err_guest_notifiers;
> > }
> >
> > -ret = vhost_dev_set_inflight(>dev, vsc->inflight);
> > -if (ret < 0) {
> > -error_report("Error set inflight: %d", -ret);
> > -goto err_guest_notifiers;
> > +if (vsc->inflight) {
> > +if (!vsc->inflight->addr) {
> > +ret = vhost_dev_get_inflight(>dev,
> > +vs->conf.virtqueue_size,
> > +vsc->inflight);
> > +if (ret < 0) {
> > +error_report("Error getting inflight: %d", -ret);
> > +goto err_guest_notifiers;
> > +}
> > +}
> > +
> > +ret = vhost_dev_set_inflight(>dev, vsc->inflight);
> > +if (ret < 0) {
> > +error_report("Error setting inflight: %d", -ret);
> > +goto err_guest_notifiers;
> > +}
> > }
> >
> > ret = vhost_dev_start(>dev, vdev, true);
> > @@ -85,9 +93,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
> > return ret;
> >
> > err_guest_notifiers:
> > -g_free(vsc->inflight);
> > -vsc->inflight = NULL;
> > -
> > k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
> > err_host_notifiers:
> > vhost_dev_disable_notifiers(>dev, vdev);
> > @@ -111,12 +116,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
> > }
> > assert(ret >= 0);
> >
> > -if (vsc->inflight) {
> > -vhost_dev_free_inflight(vsc->inflight);
> > -g_free(vsc->inflight);
> > -vsc->inflight = NULL;
> > -}
> > -
> > vhost_dev_disable_notifiers(>dev, vdev);
> > }
> >
> > --
> > 2.41.0
> >
>
Re: [PATCH v6 3/5] vhost-user-scsi: support reconnect to backend
Sorry, the reply is late due to being on vacation for half a month.
On Fri, Sep 29, 2023 at 8:55 AM Raphael Norwitz
wrote:
>
> One comment on the logging stuff in vhost-scsi. As far as I can tell the
> logging in vhost-user-scsi looks good.
>
> Markus - does this look better to you? Otherwise do you think we should also
> fix up the vhost-user-blk realize function?
>
> > On Sep 22, 2023, at 7:46 AM, Li Feng wrote:
> >
> > If the backend crashes and restarts, the device is broken.
> > This patch adds reconnect for vhost-user-scsi.
> >
> > This patch also improves the error messages, and reports some silent errors.
> >
> > Tested with spdk backend.
> >
> > Signed-off-by: Li Feng
> > ---
> > hw/scsi/vhost-scsi-common.c | 16 +-
> > hw/scsi/vhost-scsi.c | 5 +-
> > hw/scsi/vhost-user-scsi.c | 204 +++---
> > include/hw/virtio/vhost-scsi-common.h | 2 +-
> > include/hw/virtio/vhost-user-scsi.h | 4 +
> > 5 files changed, 201 insertions(+), 30 deletions(-)
> >
> > diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
> > index a61cd0e907..4c8637045d 100644
> > --- a/hw/scsi/vhost-scsi-common.c
> > +++ b/hw/scsi/vhost-scsi-common.c
> > @@ -16,6 +16,7 @@
> > */
> >
> > #include "qemu/osdep.h"
> > +#include "qapi/error.h"
> > #include "qemu/error-report.h"
> > #include "qemu/module.h"
> > #include "hw/virtio/vhost.h"
> > @@ -25,7 +26,7 @@
> > #include "hw/virtio/virtio-access.h"
> > #include "hw/fw-path-provider.h"
> >
> > -int vhost_scsi_common_start(VHostSCSICommon *vsc)
> > +int vhost_scsi_common_start(VHostSCSICommon *vsc, Error **errp)
> > {
> > int ret, i;
> > VirtIODevice *vdev = VIRTIO_DEVICE(vsc);
> > @@ -35,18 +36,19 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
> > VirtIOSCSICommon *vs = (VirtIOSCSICommon *)vsc;
> >
> > if (!k->set_guest_notifiers) {
> > -error_report("binding does not support guest notifiers");
> > +error_setg(errp, "binding does not support guest notifiers");
> > return -ENOSYS;
> > }
> >
> > ret = vhost_dev_enable_notifiers(>dev, vdev);
> > if (ret < 0) {
> > +error_setg_errno(errp, -ret, "Error enabling host notifiers");
> > return ret;
> > }
> >
> > ret = k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, true);
> > if (ret < 0) {
> > -error_report("Error binding guest notifier");
> > +error_setg_errno(errp, -ret, "Error binding guest notifier");
> > goto err_host_notifiers;
> > }
> >
> > @@ -54,7 +56,7 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
> >
> > ret = vhost_dev_prepare_inflight(>dev, vdev);
> > if (ret < 0) {
> > -error_report("Error setting inflight format: %d", -ret);
> > +error_setg_errno(errp, -ret, "Error setting inflight format");
> > goto err_guest_notifiers;
> > }
> >
> > @@ -64,21 +66,21 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
> > vs->conf.virtqueue_size,
> > vsc->inflight);
> > if (ret < 0) {
> > -error_report("Error getting inflight: %d", -ret);
> > +error_setg_errno(errp, -ret, "Error getting inflight");
> > goto err_guest_notifiers;
> > }
> > }
> >
> > ret = vhost_dev_set_inflight(>dev, vsc->inflight);
> > if (ret < 0) {
> > -error_report("Error setting inflight: %d", -ret);
> > +error_setg_errno(errp, -ret, "Error setting inflight");
> > goto err_guest_notifiers;
> > }
> > }
> >
> > ret = vhost_dev_start(>dev, vdev, true);
> > if (ret < 0) {
> > -error_report("Error start vhost dev");
> > +error_setg_errno(errp, -ret, "Error starting vhost dev");
> > goto err_guest_notifiers;
> > }
> >
> > diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c
> > index 443f67daa4..01a3ab4277 100644
> > --- a/hw/scsi/vhost-scsi.c
> > +++ b/hw/scsi/vhost-scsi.c
> > @@ -75,6 +75,7 @@ static in
[PATCH v6 0/5] Implement reconnect for vhost-user-scsi
Changes for v6: - [PATCH] vhost-user: fix lost reconnect - Fix missing assign event_cb. Changes for v5: - No logic has been changed, just move part of the code from patch 4 to patch 5. Changes for v4: - Merge https://lore.kernel.org/all/20230830045722.611224-1-fen...@smartx.com/ to this series. - Add ERRP_GUARD in vhost_user_scsi_realize; - Reword the commit messages. Changes for v3: - Split the vhost_user_scsi_handle_output to a separate patch; - Move the started_vu from vhost scsi common header to vhost-user-scsi header; - Fix a log print error; Changes for v2: - Split the v1 patch to small separate patchset; - New patch for fixing fd leak, which has sent to reviewers in another mail; - Implement the `vhost_user_scsi_handle_output`; - Add the started_vu safe check; - Fix error handler; - Check the inflight before set/get inflight fd. Li Feng (5): vhost-user-common: send get_inflight_fd once vhost: move and rename the conn retry times vhost-user-scsi: support reconnect to backend vhost-user-scsi: start vhost when guest kicks vhost-user: fix lost reconnect hw/block/vhost-user-blk.c | 6 +- hw/scsi/vhost-scsi-common.c | 47 ++--- hw/scsi/vhost-scsi.c | 5 +- hw/scsi/vhost-user-scsi.c | 253 +++--- hw/virtio/vhost-user-gpio.c | 5 +- hw/virtio/vhost-user.c| 10 +- include/hw/virtio/vhost-scsi-common.h | 2 +- include/hw/virtio/vhost-user-scsi.h | 4 + include/hw/virtio/vhost-user.h| 3 +- include/hw/virtio/vhost.h | 2 + 10 files changed, 277 insertions(+), 60 deletions(-) -- 2.41.0
[PATCH v6 4/5] vhost-user-scsi: start vhost when guest kicks
Let's keep the same behavior as vhost-user-blk.
Some old guests kick virtqueue before setting VIRTIO_CONFIG_S_DRIVER_OK.
Signed-off-by: Li Feng
---
hw/scsi/vhost-user-scsi.c | 48 +++
1 file changed, 44 insertions(+), 4 deletions(-)
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index dc109154ad..53a62c3170 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -115,8 +115,48 @@ static void vhost_user_scsi_reset(VirtIODevice *vdev)
}
}
-static void vhost_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+static void vhost_user_scsi_handle_output(VirtIODevice *vdev, VirtQueue *vq)
{
+VHostUserSCSI *s = (VHostUserSCSI *)vdev;
+DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+
+Error *local_err = NULL;
+int i, ret;
+
+if (!vdev->start_on_kick) {
+return;
+}
+
+if (!s->connected) {
+return;
+}
+
+if (vhost_dev_is_started(>dev)) {
+return;
+}
+
+/*
+ * Some guests kick before setting VIRTIO_CONFIG_S_DRIVER_OK so start
+ * vhost here instead of waiting for .set_status().
+ */
+ret = vhost_user_scsi_start(s, _err);
+if (ret < 0) {
+error_reportf_err(local_err, "vhost-user-scsi: vhost start failed: ");
+qemu_chr_fe_disconnect(>conf.chardev);
+return;
+}
+
+/* Kick right away to begin processing requests already in vring */
+for (i = 0; i < vsc->dev.nvqs; i++) {
+VirtQueue *kick_vq = virtio_get_queue(vdev, i);
+
+if (!virtio_queue_get_desc_addr(vdev, i)) {
+continue;
+}
+event_notifier_set(virtio_queue_get_host_notifier(kick_vq));
+}
}
static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
@@ -246,9 +286,9 @@ static void vhost_user_scsi_realize(DeviceState *dev, Error
**errp)
return;
}
-virtio_scsi_common_realize(dev, vhost_dummy_handle_output,
- vhost_dummy_handle_output,
- vhost_dummy_handle_output, );
+virtio_scsi_common_realize(dev, vhost_user_scsi_handle_output,
+ vhost_user_scsi_handle_output,
+ vhost_user_scsi_handle_output, );
if (err != NULL) {
error_propagate(errp, err);
return;
--
2.41.0
[PATCH v6 5/5] vhost-user: fix lost reconnect
When the vhost-user is reconnecting to the backend, and if the vhost-user fails
at the get_features in vhost_dev_init(), then the reconnect will fail
and it will not be retriggered forever.
The reason is:
When the vhost-user fails at get_features, the vhost_dev_cleanup will be called
immediately.
vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
The reconnect path is:
vhost_user_blk_event
vhost_user_async_close(.. vhost_user_blk_disconnect ..)
qemu_chr_fe_set_handlers <- clear the notifier callback
schedule vhost_user_async_close_bh
The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
called, then the event fd callback will not be reinstalled.
All vhost-user devices have this issue, including vhost-user-blk/scsi.
With this patch, if the vdev->vdev is null, the fd callback will still
be reinstalled.
Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 2 +-
hw/scsi/vhost-user-scsi.c | 3 ++-
hw/virtio/vhost-user-gpio.c| 2 +-
hw/virtio/vhost-user.c | 10 --
include/hw/virtio/vhost-user.h | 3 ++-
5 files changed, 14 insertions(+), 6 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 3c69fa47d5..95c758200d 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -391,7 +391,7 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent
event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >dev,
- vhost_user_blk_disconnect);
+ vhost_user_blk_disconnect,
vhost_user_blk_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 53a62c3170..0effbb4787 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -238,7 +238,8 @@ static void vhost_user_scsi_event(void *opaque,
QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >conf.chardev, >dev,
- vhost_user_scsi_disconnect);
+ vhost_user_scsi_disconnect,
+ vhost_user_scsi_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index d9979aa5db..04c2cc79f4 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -283,7 +283,7 @@ static void vu_gpio_event(void *opaque, QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >vhost_dev,
- vu_gpio_disconnect);
+ vu_gpio_disconnect, vu_gpio_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index 8dcf049d42..7344f57ba7 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2643,6 +2643,7 @@ typedef struct {
DeviceState *dev;
CharBackend *cd;
struct vhost_dev *vhost;
+IOEventHandler *event_cb;
} VhostAsyncCallback;
static void vhost_user_async_close_bh(void *opaque)
@@ -2657,7 +2658,10 @@ static void vhost_user_async_close_bh(void *opaque)
*/
if (vhost->vdev) {
data->cb(data->dev);
-}
+} else if (data->event_cb) {
+qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
+ NULL, data->dev, NULL, true);
+ }
g_free(data);
}
@@ -2669,7 +2673,8 @@ static void vhost_user_async_close_bh(void *opaque)
*/
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb)
+vu_async_close_fn cb,
+IOEventHandler *event_cb)
{
if (!runstate_check(RUN_STATE_SHUTDOWN)) {
/*
@@ -2685,6 +2690,7 @@ void vhost_user_async_close(DeviceState *d,
data->dev = d;
data->cd = chardev;
data->vhost = vhost;
+data->event_cb = event_cb;
/* Disable any further notifications on the chardev */
qemu_chr_fe_set_handlers(chardev,
diff --git a/include/hw/virtio/vhost-user.h b/include/hw/virtio/vhost-user.h
index 191216a74f..649e9dd54f 100644
--- a/include/hw/virtio/vhost-user.h
+++ b/include/hw/virtio/vhost-user.h
@@ -84,6 +84,7 @@ typedef void (*vu_async_close_fn)(DeviceState *cb);
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, stru
[PATCH v6 3/5] vhost-user-scsi: support reconnect to backend
If the backend crashes and restarts, the device is broken.
This patch adds reconnect for vhost-user-scsi.
This patch also improves the error messages, and reports some silent errors.
Tested with spdk backend.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 16 +-
hw/scsi/vhost-scsi.c | 5 +-
hw/scsi/vhost-user-scsi.c | 204 +++---
include/hw/virtio/vhost-scsi-common.h | 2 +-
include/hw/virtio/vhost-user-scsi.h | 4 +
5 files changed, 201 insertions(+), 30 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a61cd0e907..4c8637045d 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -16,6 +16,7 @@
*/
#include "qemu/osdep.h"
+#include "qapi/error.h"
#include "qemu/error-report.h"
#include "qemu/module.h"
#include "hw/virtio/vhost.h"
@@ -25,7 +26,7 @@
#include "hw/virtio/virtio-access.h"
#include "hw/fw-path-provider.h"
-int vhost_scsi_common_start(VHostSCSICommon *vsc)
+int vhost_scsi_common_start(VHostSCSICommon *vsc, Error **errp)
{
int ret, i;
VirtIODevice *vdev = VIRTIO_DEVICE(vsc);
@@ -35,18 +36,19 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
VirtIOSCSICommon *vs = (VirtIOSCSICommon *)vsc;
if (!k->set_guest_notifiers) {
-error_report("binding does not support guest notifiers");
+error_setg(errp, "binding does not support guest notifiers");
return -ENOSYS;
}
ret = vhost_dev_enable_notifiers(>dev, vdev);
if (ret < 0) {
+error_setg_errno(errp, -ret, "Error enabling host notifiers");
return ret;
}
ret = k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, true);
if (ret < 0) {
-error_report("Error binding guest notifier");
+error_setg_errno(errp, -ret, "Error binding guest notifier");
goto err_host_notifiers;
}
@@ -54,7 +56,7 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error setting inflight format: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight format");
goto err_guest_notifiers;
}
@@ -64,21 +66,21 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vs->conf.virtqueue_size,
vsc->inflight);
if (ret < 0) {
-error_report("Error getting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error getting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_set_inflight(>dev, vsc->inflight);
if (ret < 0) {
-error_report("Error setting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_start(>dev, vdev, true);
if (ret < 0) {
-error_report("Error start vhost dev");
+error_setg_errno(errp, -ret, "Error starting vhost dev");
goto err_guest_notifiers;
}
diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c
index 443f67daa4..01a3ab4277 100644
--- a/hw/scsi/vhost-scsi.c
+++ b/hw/scsi/vhost-scsi.c
@@ -75,6 +75,7 @@ static int vhost_scsi_start(VHostSCSI *s)
int ret, abi_version;
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
const VhostOps *vhost_ops = vsc->dev.vhost_ops;
+Error *local_err = NULL;
ret = vhost_ops->vhost_scsi_get_abi_version(>dev, _version);
if (ret < 0) {
@@ -88,14 +89,14 @@ static int vhost_scsi_start(VHostSCSI *s)
return -ENOSYS;
}
-ret = vhost_scsi_common_start(vsc);
+ret = vhost_scsi_common_start(vsc, _err);
if (ret < 0) {
return ret;
}
ret = vhost_scsi_set_endpoint(s);
if (ret < 0) {
-error_report("Error setting vhost-scsi endpoint");
+error_reportf_err(local_err, "Error setting vhost-scsi endpoint");
vhost_scsi_common_stop(vsc);
}
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index ee99b19e7a..dc109154ad 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -43,26 +43,56 @@ enum VhostUserProtocolFeature {
VHOST_USER_PROTOCOL_F_RESET_DEVICE = 13,
};
+static int vhost_user_scsi_start(VHostUserSCSI *s, Error **errp)
+{
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+int ret;
+
+ret = vhost_scsi_common_start(vsc, errp);
+s->started_vu = (ret < 0 ? false : true);
+
+return ret;
+}
+
+static void vhost_user_scsi_stop(VHostUserSCSI *s)
+{
+
[PATCH v6 2/5] vhost: move and rename the conn retry times
Multiple devices need this macro, move it to a common header.
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/block/vhost-user-blk.c | 4 +---
hw/virtio/vhost-user-gpio.c | 3 +--
include/hw/virtio/vhost.h | 2 ++
3 files changed, 4 insertions(+), 5 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index eecf3f7a81..3c69fa47d5 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -32,8 +32,6 @@
#include "sysemu/sysemu.h"
#include "sysemu/runstate.h"
-#define REALIZE_CONNECTION_RETRIES 3
-
static const int user_feature_bits[] = {
VIRTIO_BLK_F_SIZE_MAX,
VIRTIO_BLK_F_SEG_MAX,
@@ -482,7 +480,7 @@ static void vhost_user_blk_device_realize(DeviceState *dev,
Error **errp)
s->inflight = g_new0(struct vhost_inflight, 1);
s->vhost_vqs = g_new0(struct vhost_virtqueue, s->num_queues);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
assert(!*errp);
do {
if (*errp) {
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index 3b013f2d0f..d9979aa5db 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -15,7 +15,6 @@
#include "standard-headers/linux/virtio_ids.h"
#include "trace.h"
-#define REALIZE_CONNECTION_RETRIES 3
#define VHOST_NVQS 2
/* Features required from VirtIO */
@@ -359,7 +358,7 @@ static void vu_gpio_device_realize(DeviceState *dev, Error
**errp)
qemu_chr_fe_set_handlers(>chardev, NULL, NULL, vu_gpio_event, NULL,
dev, NULL, true);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
g_assert(!*errp);
do {
if (*errp) {
diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h
index 6a173cb9fa..ca3131b1af 100644
--- a/include/hw/virtio/vhost.h
+++ b/include/hw/virtio/vhost.h
@@ -8,6 +8,8 @@
#define VHOST_F_DEVICE_IOTLB 63
#define VHOST_USER_F_PROTOCOL_FEATURES 30
+#define VU_REALIZE_CONN_RETRIES 3
+
/* Generic structures common for any vhost based device. */
struct vhost_inflight {
--
2.41.0
[PATCH v6 1/5] vhost-user-common: send get_inflight_fd once
Currently the get_inflight_fd will be sent every time the device is started, and
the backend will allocate shared memory to save the inflight state. If the
backend finds that it receives the second get_inflight_fd, it will release the
previous shared memory, which breaks inflight working logic.
This patch is a preparation for the following patches.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 37 ++---
1 file changed, 18 insertions(+), 19 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a06f01af26..a61cd0e907 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -52,20 +52,28 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vsc->dev.acked_features = vdev->guest_features;
-assert(vsc->inflight == NULL);
-vsc->inflight = g_new0(struct vhost_inflight, 1);
-ret = vhost_dev_get_inflight(>dev,
- vs->conf.virtqueue_size,
- vsc->inflight);
+ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error get inflight: %d", -ret);
+error_report("Error setting inflight format: %d", -ret);
goto err_guest_notifiers;
}
-ret = vhost_dev_set_inflight(>dev, vsc->inflight);
-if (ret < 0) {
-error_report("Error set inflight: %d", -ret);
-goto err_guest_notifiers;
+if (vsc->inflight) {
+if (!vsc->inflight->addr) {
+ret = vhost_dev_get_inflight(>dev,
+vs->conf.virtqueue_size,
+vsc->inflight);
+if (ret < 0) {
+error_report("Error getting inflight: %d", -ret);
+goto err_guest_notifiers;
+}
+}
+
+ret = vhost_dev_set_inflight(>dev, vsc->inflight);
+if (ret < 0) {
+error_report("Error setting inflight: %d", -ret);
+goto err_guest_notifiers;
+}
}
ret = vhost_dev_start(>dev, vdev, true);
@@ -85,9 +93,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
return ret;
err_guest_notifiers:
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-
k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
err_host_notifiers:
vhost_dev_disable_notifiers(>dev, vdev);
@@ -111,12 +116,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
}
assert(ret >= 0);
-if (vsc->inflight) {
-vhost_dev_free_inflight(vsc->inflight);
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-}
-
vhost_dev_disable_notifiers(>dev, vdev);
}
--
2.41.0
[PATCH v5 5/5] vhost-user: fix lost reconnect
When the vhost-user is reconnecting to the backend, and if the vhost-user fails
at the get_features in vhost_dev_init(), then the reconnect will fail
and it will not be retriggered forever.
The reason is:
When the vhost-user fails at get_features, the vhost_dev_cleanup will be called
immediately.
vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
The reconnect path is:
vhost_user_blk_event
vhost_user_async_close(.. vhost_user_blk_disconnect ..)
qemu_chr_fe_set_handlers <- clear the notifier callback
schedule vhost_user_async_close_bh
The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
called, then the event fd callback will not be reinstalled.
All vhost-user devices have this issue, including vhost-user-blk/scsi.
With this patch, if the vdev->vdev is null, the fd callback will still
be reinstalled.
Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 2 +-
hw/scsi/vhost-user-scsi.c | 3 ++-
hw/virtio/vhost-user-gpio.c| 2 +-
hw/virtio/vhost-user.c | 9 +++--
include/hw/virtio/vhost-user.h | 3 ++-
5 files changed, 13 insertions(+), 6 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 3c69fa47d5..95c758200d 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -391,7 +391,7 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent
event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >dev,
- vhost_user_blk_disconnect);
+ vhost_user_blk_disconnect,
vhost_user_blk_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 53a62c3170..0effbb4787 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -238,7 +238,8 @@ static void vhost_user_scsi_event(void *opaque,
QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >conf.chardev, >dev,
- vhost_user_scsi_disconnect);
+ vhost_user_scsi_disconnect,
+ vhost_user_scsi_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index d9979aa5db..04c2cc79f4 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -283,7 +283,7 @@ static void vu_gpio_event(void *opaque, QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >vhost_dev,
- vu_gpio_disconnect);
+ vu_gpio_disconnect, vu_gpio_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index 8dcf049d42..12c4a41f30 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2643,6 +2643,7 @@ typedef struct {
DeviceState *dev;
CharBackend *cd;
struct vhost_dev *vhost;
+IOEventHandler *event_cb;
} VhostAsyncCallback;
static void vhost_user_async_close_bh(void *opaque)
@@ -2657,7 +2658,10 @@ static void vhost_user_async_close_bh(void *opaque)
*/
if (vhost->vdev) {
data->cb(data->dev);
-}
+} else if (data->event_cb) {
+qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
+ NULL, data->dev, NULL, true);
+ }
g_free(data);
}
@@ -2669,7 +2673,8 @@ static void vhost_user_async_close_bh(void *opaque)
*/
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb)
+vu_async_close_fn cb,
+IOEventHandler *event_cb)
{
if (!runstate_check(RUN_STATE_SHUTDOWN)) {
/*
diff --git a/include/hw/virtio/vhost-user.h b/include/hw/virtio/vhost-user.h
index 191216a74f..649e9dd54f 100644
--- a/include/hw/virtio/vhost-user.h
+++ b/include/hw/virtio/vhost-user.h
@@ -84,6 +84,7 @@ typedef void (*vu_async_close_fn)(DeviceState *cb);
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb);
+vu_async_close_fn cb,
+IOEventHandler *event_cb);
#endif
--
2.41.0
[PATCH v5 0/5] Implement reconnect for vhost-user-scsi
Ping ... Could anyone review this series and merge them? Changes for v5: - No logic has been changed, just move part of the code from patch 4 to patch 5. Changes for v4: - Merge https://lore.kernel.org/all/20230830045722.611224-1-fen...@smartx.com/ to this series. - Add ERRP_GUARD in vhost_user_scsi_realize; - Reword the commit messages. Changes for v3: - Split the vhost_user_scsi_handle_output to a separate patch; - Move the started_vu from vhost scsi common header to vhost-user-scsi header; - Fix a log print error; Changes for v2: - Split the v1 patch to small separate patchset; - New patch for fixing fd leak, which has sent to reviewers in another mail; - Implement the `vhost_user_scsi_handle_output`; - Add the started_vu safe check; - Fix error handler; - Check the inflight before set/get inflight fd. Li Feng (5): vhost-user-common: send get_inflight_fd once vhost: move and rename the conn retry times vhost-user-scsi: support reconnect to backend vhost-user-scsi: start vhost when guest kicks vhost-user: fix lost reconnect hw/block/vhost-user-blk.c | 6 +- hw/scsi/vhost-scsi-common.c | 47 ++--- hw/scsi/vhost-scsi.c | 5 +- hw/scsi/vhost-user-scsi.c | 253 +++--- hw/virtio/vhost-user-gpio.c | 5 +- hw/virtio/vhost-user.c| 9 +- include/hw/virtio/vhost-scsi-common.h | 2 +- include/hw/virtio/vhost-user-scsi.h | 4 + include/hw/virtio/vhost-user.h| 3 +- include/hw/virtio/vhost.h | 2 + 10 files changed, 276 insertions(+), 60 deletions(-) -- 2.41.0
[PATCH v5 2/5] vhost: move and rename the conn retry times
Multiple devices need this macro, move it to a common header.
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/block/vhost-user-blk.c | 4 +---
hw/virtio/vhost-user-gpio.c | 3 +--
include/hw/virtio/vhost.h | 2 ++
3 files changed, 4 insertions(+), 5 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index eecf3f7a81..3c69fa47d5 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -32,8 +32,6 @@
#include "sysemu/sysemu.h"
#include "sysemu/runstate.h"
-#define REALIZE_CONNECTION_RETRIES 3
-
static const int user_feature_bits[] = {
VIRTIO_BLK_F_SIZE_MAX,
VIRTIO_BLK_F_SEG_MAX,
@@ -482,7 +480,7 @@ static void vhost_user_blk_device_realize(DeviceState *dev,
Error **errp)
s->inflight = g_new0(struct vhost_inflight, 1);
s->vhost_vqs = g_new0(struct vhost_virtqueue, s->num_queues);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
assert(!*errp);
do {
if (*errp) {
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index 3b013f2d0f..d9979aa5db 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -15,7 +15,6 @@
#include "standard-headers/linux/virtio_ids.h"
#include "trace.h"
-#define REALIZE_CONNECTION_RETRIES 3
#define VHOST_NVQS 2
/* Features required from VirtIO */
@@ -359,7 +358,7 @@ static void vu_gpio_device_realize(DeviceState *dev, Error
**errp)
qemu_chr_fe_set_handlers(>chardev, NULL, NULL, vu_gpio_event, NULL,
dev, NULL, true);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
g_assert(!*errp);
do {
if (*errp) {
diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h
index 6a173cb9fa..ca3131b1af 100644
--- a/include/hw/virtio/vhost.h
+++ b/include/hw/virtio/vhost.h
@@ -8,6 +8,8 @@
#define VHOST_F_DEVICE_IOTLB 63
#define VHOST_USER_F_PROTOCOL_FEATURES 30
+#define VU_REALIZE_CONN_RETRIES 3
+
/* Generic structures common for any vhost based device. */
struct vhost_inflight {
--
2.41.0
[PATCH v5 1/5] vhost-user-common: send get_inflight_fd once
Currently the get_inflight_fd will be sent every time the device is started, and
the backend will allocate shared memory to save the inflight state. If the
backend finds that it receives the second get_inflight_fd, it will release the
previous shared memory, which breaks inflight working logic.
This patch is a preparation for the following patches.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 37 ++---
1 file changed, 18 insertions(+), 19 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a06f01af26..a61cd0e907 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -52,20 +52,28 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vsc->dev.acked_features = vdev->guest_features;
-assert(vsc->inflight == NULL);
-vsc->inflight = g_new0(struct vhost_inflight, 1);
-ret = vhost_dev_get_inflight(>dev,
- vs->conf.virtqueue_size,
- vsc->inflight);
+ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error get inflight: %d", -ret);
+error_report("Error setting inflight format: %d", -ret);
goto err_guest_notifiers;
}
-ret = vhost_dev_set_inflight(>dev, vsc->inflight);
-if (ret < 0) {
-error_report("Error set inflight: %d", -ret);
-goto err_guest_notifiers;
+if (vsc->inflight) {
+if (!vsc->inflight->addr) {
+ret = vhost_dev_get_inflight(>dev,
+vs->conf.virtqueue_size,
+vsc->inflight);
+if (ret < 0) {
+error_report("Error getting inflight: %d", -ret);
+goto err_guest_notifiers;
+}
+}
+
+ret = vhost_dev_set_inflight(>dev, vsc->inflight);
+if (ret < 0) {
+error_report("Error setting inflight: %d", -ret);
+goto err_guest_notifiers;
+}
}
ret = vhost_dev_start(>dev, vdev, true);
@@ -85,9 +93,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
return ret;
err_guest_notifiers:
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-
k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
err_host_notifiers:
vhost_dev_disable_notifiers(>dev, vdev);
@@ -111,12 +116,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
}
assert(ret >= 0);
-if (vsc->inflight) {
-vhost_dev_free_inflight(vsc->inflight);
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-}
-
vhost_dev_disable_notifiers(>dev, vdev);
}
--
2.41.0
[PATCH v5 3/5] vhost-user-scsi: support reconnect to backend
If the backend crashes and restarts, the device is broken.
This patch adds reconnect for vhost-user-scsi.
This patch also improves the error messages, and reports some silent errors.
Tested with spdk backend.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 16 +-
hw/scsi/vhost-scsi.c | 5 +-
hw/scsi/vhost-user-scsi.c | 204 +++---
include/hw/virtio/vhost-scsi-common.h | 2 +-
include/hw/virtio/vhost-user-scsi.h | 4 +
5 files changed, 201 insertions(+), 30 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a61cd0e907..4c8637045d 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -16,6 +16,7 @@
*/
#include "qemu/osdep.h"
+#include "qapi/error.h"
#include "qemu/error-report.h"
#include "qemu/module.h"
#include "hw/virtio/vhost.h"
@@ -25,7 +26,7 @@
#include "hw/virtio/virtio-access.h"
#include "hw/fw-path-provider.h"
-int vhost_scsi_common_start(VHostSCSICommon *vsc)
+int vhost_scsi_common_start(VHostSCSICommon *vsc, Error **errp)
{
int ret, i;
VirtIODevice *vdev = VIRTIO_DEVICE(vsc);
@@ -35,18 +36,19 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
VirtIOSCSICommon *vs = (VirtIOSCSICommon *)vsc;
if (!k->set_guest_notifiers) {
-error_report("binding does not support guest notifiers");
+error_setg(errp, "binding does not support guest notifiers");
return -ENOSYS;
}
ret = vhost_dev_enable_notifiers(>dev, vdev);
if (ret < 0) {
+error_setg_errno(errp, -ret, "Error enabling host notifiers");
return ret;
}
ret = k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, true);
if (ret < 0) {
-error_report("Error binding guest notifier");
+error_setg_errno(errp, -ret, "Error binding guest notifier");
goto err_host_notifiers;
}
@@ -54,7 +56,7 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error setting inflight format: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight format");
goto err_guest_notifiers;
}
@@ -64,21 +66,21 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vs->conf.virtqueue_size,
vsc->inflight);
if (ret < 0) {
-error_report("Error getting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error getting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_set_inflight(>dev, vsc->inflight);
if (ret < 0) {
-error_report("Error setting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_start(>dev, vdev, true);
if (ret < 0) {
-error_report("Error start vhost dev");
+error_setg_errno(errp, -ret, "Error starting vhost dev");
goto err_guest_notifiers;
}
diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c
index 443f67daa4..01a3ab4277 100644
--- a/hw/scsi/vhost-scsi.c
+++ b/hw/scsi/vhost-scsi.c
@@ -75,6 +75,7 @@ static int vhost_scsi_start(VHostSCSI *s)
int ret, abi_version;
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
const VhostOps *vhost_ops = vsc->dev.vhost_ops;
+Error *local_err = NULL;
ret = vhost_ops->vhost_scsi_get_abi_version(>dev, _version);
if (ret < 0) {
@@ -88,14 +89,14 @@ static int vhost_scsi_start(VHostSCSI *s)
return -ENOSYS;
}
-ret = vhost_scsi_common_start(vsc);
+ret = vhost_scsi_common_start(vsc, _err);
if (ret < 0) {
return ret;
}
ret = vhost_scsi_set_endpoint(s);
if (ret < 0) {
-error_report("Error setting vhost-scsi endpoint");
+error_reportf_err(local_err, "Error setting vhost-scsi endpoint");
vhost_scsi_common_stop(vsc);
}
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index ee99b19e7a..dc109154ad 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -43,26 +43,56 @@ enum VhostUserProtocolFeature {
VHOST_USER_PROTOCOL_F_RESET_DEVICE = 13,
};
+static int vhost_user_scsi_start(VHostUserSCSI *s, Error **errp)
+{
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+int ret;
+
+ret = vhost_scsi_common_start(vsc, errp);
+s->started_vu = (ret < 0 ? false : true);
+
+return ret;
+}
+
+static void vhost_user_scsi_stop(VHostUserSCSI *s)
+{
+
[PATCH v5 4/5] vhost-user-scsi: start vhost when guest kicks
Let's keep the same behavior as vhost-user-blk.
Some old guests kick virtqueue before setting VIRTIO_CONFIG_S_DRIVER_OK.
Signed-off-by: Li Feng
---
hw/scsi/vhost-user-scsi.c | 48 +++
1 file changed, 44 insertions(+), 4 deletions(-)
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index dc109154ad..53a62c3170 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -115,8 +115,48 @@ static void vhost_user_scsi_reset(VirtIODevice *vdev)
}
}
-static void vhost_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+static void vhost_user_scsi_handle_output(VirtIODevice *vdev, VirtQueue *vq)
{
+VHostUserSCSI *s = (VHostUserSCSI *)vdev;
+DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+
+Error *local_err = NULL;
+int i, ret;
+
+if (!vdev->start_on_kick) {
+return;
+}
+
+if (!s->connected) {
+return;
+}
+
+if (vhost_dev_is_started(>dev)) {
+return;
+}
+
+/*
+ * Some guests kick before setting VIRTIO_CONFIG_S_DRIVER_OK so start
+ * vhost here instead of waiting for .set_status().
+ */
+ret = vhost_user_scsi_start(s, _err);
+if (ret < 0) {
+error_reportf_err(local_err, "vhost-user-scsi: vhost start failed: ");
+qemu_chr_fe_disconnect(>conf.chardev);
+return;
+}
+
+/* Kick right away to begin processing requests already in vring */
+for (i = 0; i < vsc->dev.nvqs; i++) {
+VirtQueue *kick_vq = virtio_get_queue(vdev, i);
+
+if (!virtio_queue_get_desc_addr(vdev, i)) {
+continue;
+}
+event_notifier_set(virtio_queue_get_host_notifier(kick_vq));
+}
}
static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
@@ -246,9 +286,9 @@ static void vhost_user_scsi_realize(DeviceState *dev, Error
**errp)
return;
}
-virtio_scsi_common_realize(dev, vhost_dummy_handle_output,
- vhost_dummy_handle_output,
- vhost_dummy_handle_output, );
+virtio_scsi_common_realize(dev, vhost_user_scsi_handle_output,
+ vhost_user_scsi_handle_output,
+ vhost_user_scsi_handle_output, );
if (err != NULL) {
error_propagate(errp, err);
return;
--
2.41.0
[PATCH v4 3/5] vhost-user-scsi: support reconnect to backend
If the backend crashes and restarts, the device is broken.
This patch adds reconnect for vhost-user-scsi.
This patch also improves the error messages, and reports some silent errors.
Tested with spdk backend.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 16 +-
hw/scsi/vhost-scsi.c | 5 +-
hw/scsi/vhost-user-scsi.c | 204 +++---
include/hw/virtio/vhost-scsi-common.h | 2 +-
include/hw/virtio/vhost-user-scsi.h | 4 +
5 files changed, 201 insertions(+), 30 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a61cd0e907..4c8637045d 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -16,6 +16,7 @@
*/
#include "qemu/osdep.h"
+#include "qapi/error.h"
#include "qemu/error-report.h"
#include "qemu/module.h"
#include "hw/virtio/vhost.h"
@@ -25,7 +26,7 @@
#include "hw/virtio/virtio-access.h"
#include "hw/fw-path-provider.h"
-int vhost_scsi_common_start(VHostSCSICommon *vsc)
+int vhost_scsi_common_start(VHostSCSICommon *vsc, Error **errp)
{
int ret, i;
VirtIODevice *vdev = VIRTIO_DEVICE(vsc);
@@ -35,18 +36,19 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
VirtIOSCSICommon *vs = (VirtIOSCSICommon *)vsc;
if (!k->set_guest_notifiers) {
-error_report("binding does not support guest notifiers");
+error_setg(errp, "binding does not support guest notifiers");
return -ENOSYS;
}
ret = vhost_dev_enable_notifiers(>dev, vdev);
if (ret < 0) {
+error_setg_errno(errp, -ret, "Error enabling host notifiers");
return ret;
}
ret = k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, true);
if (ret < 0) {
-error_report("Error binding guest notifier");
+error_setg_errno(errp, -ret, "Error binding guest notifier");
goto err_host_notifiers;
}
@@ -54,7 +56,7 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error setting inflight format: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight format");
goto err_guest_notifiers;
}
@@ -64,21 +66,21 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vs->conf.virtqueue_size,
vsc->inflight);
if (ret < 0) {
-error_report("Error getting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error getting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_set_inflight(>dev, vsc->inflight);
if (ret < 0) {
-error_report("Error setting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_start(>dev, vdev, true);
if (ret < 0) {
-error_report("Error start vhost dev");
+error_setg_errno(errp, -ret, "Error starting vhost dev");
goto err_guest_notifiers;
}
diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c
index 443f67daa4..01a3ab4277 100644
--- a/hw/scsi/vhost-scsi.c
+++ b/hw/scsi/vhost-scsi.c
@@ -75,6 +75,7 @@ static int vhost_scsi_start(VHostSCSI *s)
int ret, abi_version;
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
const VhostOps *vhost_ops = vsc->dev.vhost_ops;
+Error *local_err = NULL;
ret = vhost_ops->vhost_scsi_get_abi_version(>dev, _version);
if (ret < 0) {
@@ -88,14 +89,14 @@ static int vhost_scsi_start(VHostSCSI *s)
return -ENOSYS;
}
-ret = vhost_scsi_common_start(vsc);
+ret = vhost_scsi_common_start(vsc, _err);
if (ret < 0) {
return ret;
}
ret = vhost_scsi_set_endpoint(s);
if (ret < 0) {
-error_report("Error setting vhost-scsi endpoint");
+error_reportf_err(local_err, "Error setting vhost-scsi endpoint");
vhost_scsi_common_stop(vsc);
}
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index ee99b19e7a..dc109154ad 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -43,26 +43,56 @@ enum VhostUserProtocolFeature {
VHOST_USER_PROTOCOL_F_RESET_DEVICE = 13,
};
+static int vhost_user_scsi_start(VHostUserSCSI *s, Error **errp)
+{
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+int ret;
+
+ret = vhost_scsi_common_start(vsc, errp);
+s->started_vu = (ret < 0 ? false : true);
+
+return ret;
+}
+
+static void vhost_user_scsi_stop(VHostUserSCSI *s)
+{
+
[PATCH v4 5/5] vhost-user: fix lost reconnect
When the vhost-user is reconnecting to the backend, and if the vhost-user fails
at the get_features in vhost_dev_init(), then the reconnect will fail
and it will not be retriggered forever.
The reason is:
When the vhost-user fails at get_features, the vhost_dev_cleanup will be called
immediately.
vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
The reconnect path is:
vhost_user_blk_event
vhost_user_async_close(.. vhost_user_blk_disconnect ..)
qemu_chr_fe_set_handlers <- clear the notifier callback
schedule vhost_user_async_close_bh
The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
called, then the event fd callback will not be reinstalled.
All vhost-user devices have this issue, including vhost-user-blk/scsi.
With this patch, if the vdev->vdev is null, the fd callback will still
be reinstalled.
Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 2 +-
hw/virtio/vhost-user-gpio.c| 2 +-
hw/virtio/vhost-user.c | 9 +++--
include/hw/virtio/vhost-user.h | 3 ++-
4 files changed, 11 insertions(+), 5 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 3c69fa47d5..95c758200d 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -391,7 +391,7 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent
event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >dev,
- vhost_user_blk_disconnect);
+ vhost_user_blk_disconnect,
vhost_user_blk_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index d9979aa5db..04c2cc79f4 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -283,7 +283,7 @@ static void vu_gpio_event(void *opaque, QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >vhost_dev,
- vu_gpio_disconnect);
+ vu_gpio_disconnect, vu_gpio_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index 8dcf049d42..12c4a41f30 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2643,6 +2643,7 @@ typedef struct {
DeviceState *dev;
CharBackend *cd;
struct vhost_dev *vhost;
+IOEventHandler *event_cb;
} VhostAsyncCallback;
static void vhost_user_async_close_bh(void *opaque)
@@ -2657,7 +2658,10 @@ static void vhost_user_async_close_bh(void *opaque)
*/
if (vhost->vdev) {
data->cb(data->dev);
-}
+} else if (data->event_cb) {
+qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
+ NULL, data->dev, NULL, true);
+ }
g_free(data);
}
@@ -2669,7 +2673,8 @@ static void vhost_user_async_close_bh(void *opaque)
*/
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb)
+vu_async_close_fn cb,
+IOEventHandler *event_cb)
{
if (!runstate_check(RUN_STATE_SHUTDOWN)) {
/*
diff --git a/include/hw/virtio/vhost-user.h b/include/hw/virtio/vhost-user.h
index 191216a74f..649e9dd54f 100644
--- a/include/hw/virtio/vhost-user.h
+++ b/include/hw/virtio/vhost-user.h
@@ -84,6 +84,7 @@ typedef void (*vu_async_close_fn)(DeviceState *cb);
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb);
+vu_async_close_fn cb,
+IOEventHandler *event_cb);
#endif
--
2.41.0
[PATCH v4 1/5] vhost-user-common: send get_inflight_fd once
Currently the get_inflight_fd will be sent every time the device is started, and
the backend will allocate shared memory to save the inflight state. If the
backend finds that it receives the second get_inflight_fd, it will release the
previous shared memory, which breaks inflight working logic.
This patch is a preparation for the following patches.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 37 ++---
1 file changed, 18 insertions(+), 19 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a06f01af26..a61cd0e907 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -52,20 +52,28 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vsc->dev.acked_features = vdev->guest_features;
-assert(vsc->inflight == NULL);
-vsc->inflight = g_new0(struct vhost_inflight, 1);
-ret = vhost_dev_get_inflight(>dev,
- vs->conf.virtqueue_size,
- vsc->inflight);
+ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error get inflight: %d", -ret);
+error_report("Error setting inflight format: %d", -ret);
goto err_guest_notifiers;
}
-ret = vhost_dev_set_inflight(>dev, vsc->inflight);
-if (ret < 0) {
-error_report("Error set inflight: %d", -ret);
-goto err_guest_notifiers;
+if (vsc->inflight) {
+if (!vsc->inflight->addr) {
+ret = vhost_dev_get_inflight(>dev,
+vs->conf.virtqueue_size,
+vsc->inflight);
+if (ret < 0) {
+error_report("Error getting inflight: %d", -ret);
+goto err_guest_notifiers;
+}
+}
+
+ret = vhost_dev_set_inflight(>dev, vsc->inflight);
+if (ret < 0) {
+error_report("Error setting inflight: %d", -ret);
+goto err_guest_notifiers;
+}
}
ret = vhost_dev_start(>dev, vdev, true);
@@ -85,9 +93,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
return ret;
err_guest_notifiers:
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-
k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
err_host_notifiers:
vhost_dev_disable_notifiers(>dev, vdev);
@@ -111,12 +116,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
}
assert(ret >= 0);
-if (vsc->inflight) {
-vhost_dev_free_inflight(vsc->inflight);
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-}
-
vhost_dev_disable_notifiers(>dev, vdev);
}
--
2.41.0
[PATCH v4 0/5] Implement reconnect for vhost-user-scsi
This patchset adds reconnect support for vhost-user-scsi. At the same time, improve the error messages and silent errors are now reported. And fix a lost reconnect issue for all vhost-user backend. Changes for v4: - Merge https://lore.kernel.org/all/20230830045722.611224-1-fen...@smartx.com/ to this series. - Add ERRP_GUARD in vhost_user_scsi_realize; - Reword the commit messages. Changes for v3: - Split the vhost_user_scsi_handle_output to a separate patch; - Move the started_vu from vhost scsi common header to vhost-user-scsi header; - Fix a log print error; Changes for v2: - Split the v1 patch to small separate patchset; - New patch for fixing fd leak, which has sent to reviewers in another mail; - Implement the `vhost_user_scsi_handle_output`; - Add the started_vu safe check; - Fix error handler; - Check the inflight before set/get inflight fd. Li Feng (5): vhost-user-common: send get_inflight_fd once vhost: move and rename the conn retry times vhost-user-scsi: support reconnect to backend vhost-user-scsi: start vhost when guest kicks vhost-user: fix lost reconnect hw/block/vhost-user-blk.c | 6 +- hw/scsi/vhost-scsi-common.c | 47 ++--- hw/scsi/vhost-scsi.c | 5 +- hw/scsi/vhost-user-scsi.c | 253 +++--- hw/virtio/vhost-user-gpio.c | 5 +- hw/virtio/vhost-user.c| 9 +- include/hw/virtio/vhost-scsi-common.h | 2 +- include/hw/virtio/vhost-user-scsi.h | 4 + include/hw/virtio/vhost-user.h| 3 +- include/hw/virtio/vhost.h | 2 + 10 files changed, 276 insertions(+), 60 deletions(-) -- 2.41.0
[PATCH v4 4/5] vhost-user-scsi: start vhost when guest kicks
Let's keep the same behavior as vhost-user-blk.
Some old guests kick virtqueue before setting VIRTIO_CONFIG_S_DRIVER_OK.
Signed-off-by: Li Feng
---
hw/scsi/vhost-user-scsi.c | 51 +++
1 file changed, 46 insertions(+), 5 deletions(-)
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index dc109154ad..0effbb4787 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -115,8 +115,48 @@ static void vhost_user_scsi_reset(VirtIODevice *vdev)
}
}
-static void vhost_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+static void vhost_user_scsi_handle_output(VirtIODevice *vdev, VirtQueue *vq)
{
+VHostUserSCSI *s = (VHostUserSCSI *)vdev;
+DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+
+Error *local_err = NULL;
+int i, ret;
+
+if (!vdev->start_on_kick) {
+return;
+}
+
+if (!s->connected) {
+return;
+}
+
+if (vhost_dev_is_started(>dev)) {
+return;
+}
+
+/*
+ * Some guests kick before setting VIRTIO_CONFIG_S_DRIVER_OK so start
+ * vhost here instead of waiting for .set_status().
+ */
+ret = vhost_user_scsi_start(s, _err);
+if (ret < 0) {
+error_reportf_err(local_err, "vhost-user-scsi: vhost start failed: ");
+qemu_chr_fe_disconnect(>conf.chardev);
+return;
+}
+
+/* Kick right away to begin processing requests already in vring */
+for (i = 0; i < vsc->dev.nvqs; i++) {
+VirtQueue *kick_vq = virtio_get_queue(vdev, i);
+
+if (!virtio_queue_get_desc_addr(vdev, i)) {
+continue;
+}
+event_notifier_set(virtio_queue_get_host_notifier(kick_vq));
+}
}
static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
@@ -198,7 +238,8 @@ static void vhost_user_scsi_event(void *opaque,
QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >conf.chardev, >dev,
- vhost_user_scsi_disconnect);
+ vhost_user_scsi_disconnect,
+ vhost_user_scsi_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
@@ -246,9 +287,9 @@ static void vhost_user_scsi_realize(DeviceState *dev, Error
**errp)
return;
}
-virtio_scsi_common_realize(dev, vhost_dummy_handle_output,
- vhost_dummy_handle_output,
- vhost_dummy_handle_output, );
+virtio_scsi_common_realize(dev, vhost_user_scsi_handle_output,
+ vhost_user_scsi_handle_output,
+ vhost_user_scsi_handle_output, );
if (err != NULL) {
error_propagate(errp, err);
return;
--
2.41.0
[PATCH v4 2/5] vhost: move and rename the conn retry times
Multiple devices need this macro, move it to a common header.
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/block/vhost-user-blk.c | 4 +---
hw/virtio/vhost-user-gpio.c | 3 +--
include/hw/virtio/vhost.h | 2 ++
3 files changed, 4 insertions(+), 5 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index eecf3f7a81..3c69fa47d5 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -32,8 +32,6 @@
#include "sysemu/sysemu.h"
#include "sysemu/runstate.h"
-#define REALIZE_CONNECTION_RETRIES 3
-
static const int user_feature_bits[] = {
VIRTIO_BLK_F_SIZE_MAX,
VIRTIO_BLK_F_SEG_MAX,
@@ -482,7 +480,7 @@ static void vhost_user_blk_device_realize(DeviceState *dev,
Error **errp)
s->inflight = g_new0(struct vhost_inflight, 1);
s->vhost_vqs = g_new0(struct vhost_virtqueue, s->num_queues);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
assert(!*errp);
do {
if (*errp) {
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index 3b013f2d0f..d9979aa5db 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -15,7 +15,6 @@
#include "standard-headers/linux/virtio_ids.h"
#include "trace.h"
-#define REALIZE_CONNECTION_RETRIES 3
#define VHOST_NVQS 2
/* Features required from VirtIO */
@@ -359,7 +358,7 @@ static void vu_gpio_device_realize(DeviceState *dev, Error
**errp)
qemu_chr_fe_set_handlers(>chardev, NULL, NULL, vu_gpio_event, NULL,
dev, NULL, true);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
g_assert(!*errp);
do {
if (*errp) {
diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h
index 6a173cb9fa..ca3131b1af 100644
--- a/include/hw/virtio/vhost.h
+++ b/include/hw/virtio/vhost.h
@@ -8,6 +8,8 @@
#define VHOST_F_DEVICE_IOTLB 63
#define VHOST_USER_F_PROTOCOL_FEATURES 30
+#define VU_REALIZE_CONN_RETRIES 3
+
/* Generic structures common for any vhost based device. */
struct vhost_inflight {
--
2.41.0
Re: [PATCH v3 2/2] vhost: Add Error parameter to vhost_scsi_common_start()
> On 1 Sep 2023, at 8:00 PM, Markus Armbruster wrote:
>
> Li Feng mailto:fen...@smartx.com>> writes:
>
>> Add a Error parameter to report the real error, like vhost-user-blk.
>>
>> Signed-off-by: Li Feng
>> ---
>> hw/scsi/vhost-scsi-common.c | 16 +---
>> hw/scsi/vhost-scsi.c | 5 +++--
>> hw/scsi/vhost-user-scsi.c | 14 --
>> include/hw/virtio/vhost-scsi-common.h | 2 +-
>> 4 files changed, 21 insertions(+), 16 deletions(-)
>>
>> diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
>> index a61cd0e907..4c8637045d 100644
>> --- a/hw/scsi/vhost-scsi-common.c
>> +++ b/hw/scsi/vhost-scsi-common.c
>> @@ -16,6 +16,7 @@
>> */
>>
>> #include "qemu/osdep.h"
>> +#include "qapi/error.h"
>> #include "qemu/error-report.h"
>> #include "qemu/module.h"
>> #include "hw/virtio/vhost.h"
>> @@ -25,7 +26,7 @@
>> #include "hw/virtio/virtio-access.h"
>> #include "hw/fw-path-provider.h"
>>
>> -int vhost_scsi_common_start(VHostSCSICommon *vsc)
>> +int vhost_scsi_common_start(VHostSCSICommon *vsc, Error **errp)
>> {
>> int ret, i;
>> VirtIODevice *vdev = VIRTIO_DEVICE(vsc);
>> @@ -35,18 +36,19 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>> VirtIOSCSICommon *vs = (VirtIOSCSICommon *)vsc;
>>
>> if (!k->set_guest_notifiers) {
>> -error_report("binding does not support guest notifiers");
>> +error_setg(errp, "binding does not support guest notifiers");
>> return -ENOSYS;
>> }
>>
>> ret = vhost_dev_enable_notifiers(>dev, vdev);
>> if (ret < 0) {
>> +error_setg_errno(errp, -ret, "Error enabling host notifiers");
>
> Looks like the error is silent before your patch. Correct?
Yes, before this patch, it’s a silent error.
>
>> return ret;
>> }
>>
>> ret = k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, true);
>> if (ret < 0) {
>> -error_report("Error binding guest notifier");
>> +error_setg_errno(errp, -ret, "Error binding guest notifier");
>
> Error message now provides more detail.
Yes.
>
>> goto err_host_notifiers;
>> }
>>
>> @@ -54,7 +56,7 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>>
>> ret = vhost_dev_prepare_inflight(>dev, vdev);
>> if (ret < 0) {
>> -error_report("Error setting inflight format: %d", -ret);
>> +error_setg_errno(errp, -ret, "Error setting inflight format");
>
> Error message now shows errno in human-readable form.
Yes.
>
>> goto err_guest_notifiers;
>> }
>>
>> @@ -64,21 +66,21 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>> vs->conf.virtqueue_size,
>> vsc->inflight);
>> if (ret < 0) {
>> -error_report("Error getting inflight: %d", -ret);
>> +error_setg_errno(errp, -ret, "Error getting inflight");
>
> Likewise.
Yes.
>
>> goto err_guest_notifiers;
>> }
>> }
>>
>> ret = vhost_dev_set_inflight(>dev, vsc->inflight);
>> if (ret < 0) {
>> -error_report("Error setting inflight: %d", -ret);
>> +error_setg_errno(errp, -ret, "Error setting inflight");
>
> Likewise.
Yes.
>
>> goto err_guest_notifiers;
>> }
>> }
>>
>> ret = vhost_dev_start(>dev, vdev, true);
>> if (ret < 0) {
>> -error_report("Error start vhost dev");
>> +error_setg_errno(errp, -ret, "Error starting vhost dev");
>
> Likewise.
Yes.
>
>> goto err_guest_notifiers;
>> }
>>
>> diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c
>> index 443f67daa4..01a3ab4277 100644
>> --- a/hw/scsi/vhost-scsi.c
>> +++ b/hw/scsi/vhost-scsi.c
>> @@ -75,6 +75,7 @@ static int vhost_scsi_start(VHostSCSI *s)
>> int ret, abi_version;
>> VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
>> const VhostOps *vhost_ops = vsc->dev.vhost_ops;
>> +Error *local_err = NULL;
>>
>> ret = vhost_ops->vhost_scsi_get_abi_versio
Re: [PATCH v3 4/5] vhost-user-scsi: support reconnect to backend
> On 1 Sep 2023, at 8:00 PM, Markus Armbruster wrote:
>
> Li Feng mailto:fen...@smartx.com>> writes:
>
>> If the backend crashes and restarts, the device is broken.
>> This patch adds reconnect for vhost-user-scsi.
>>
>> Tested with spdk backend.
>>
>> Signed-off-by: Li Feng
>> ---
>> hw/scsi/vhost-user-scsi.c | 199 +---
>> include/hw/virtio/vhost-user-scsi.h | 4 +
>> 2 files changed, 184 insertions(+), 19 deletions(-)
>>
>> diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
>> index ee99b19e7a..5bf012461b 100644
>> --- a/hw/scsi/vhost-user-scsi.c
>> +++ b/hw/scsi/vhost-user-scsi.c
>> @@ -43,26 +43,54 @@ enum VhostUserProtocolFeature {
>> VHOST_USER_PROTOCOL_F_RESET_DEVICE = 13,
>> };
>>
>> +static int vhost_user_scsi_start(VHostUserSCSI *s)
>> +{
>> +VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
>> +int ret;
>> +
>> +ret = vhost_scsi_common_start(vsc);
>> +s->started_vu = (ret < 0 ? false : true);
>> +
>> +return ret;
>> +}
>> +
>> +static void vhost_user_scsi_stop(VHostUserSCSI *s)
>> +{
>> +VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
>> +
>> +if (!s->started_vu) {
>> +return;
>> +}
>> +s->started_vu = false;
>> +
>> +vhost_scsi_common_stop(vsc);
>> +}
>> +
>> static void vhost_user_scsi_set_status(VirtIODevice *vdev, uint8_t status)
>> {
>> VHostUserSCSI *s = (VHostUserSCSI *)vdev;
>> +DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
>> VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
>> -bool start = (status & VIRTIO_CONFIG_S_DRIVER_OK) && vdev->vm_running;
>> +VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
>> +bool should_start = virtio_device_should_start(vdev, status);
>> +int ret;
>>
>> -if (vhost_dev_is_started(>dev) == start) {
>> +if (!s->connected) {
>> return;
>> }
>>
>> -if (start) {
>> -int ret;
>> +if (vhost_dev_is_started(>dev) == should_start) {
>> +return;
>> +}
>>
>> -ret = vhost_scsi_common_start(vsc);
>> +if (should_start) {
>> +ret = vhost_user_scsi_start(s);
>> if (ret < 0) {
>> error_report("unable to start vhost-user-scsi: %s",
>> strerror(-ret));
>> -exit(1);
>> +qemu_chr_fe_disconnect(>conf.chardev);
>> }
>> } else {
>> -vhost_scsi_common_stop(vsc);
>> +vhost_user_scsi_stop(s);
>> }
>> }
>>
>> @@ -89,14 +117,126 @@ static void vhost_dummy_handle_output(VirtIODevice
>> *vdev, VirtQueue *vq)
>> {
>> }
>>
>> +static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
>> +{
>> +VirtIODevice *vdev = VIRTIO_DEVICE(dev);
>> +VHostUserSCSI *s = VHOST_USER_SCSI(vdev);
>> +VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
>> +VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
>> +int ret = 0;
>> +
>> +if (s->connected) {
>> +return 0;
>> +}
>> +s->connected = true;
>> +
>> +vsc->dev.num_queues = vs->conf.num_queues;
>> +vsc->dev.nvqs = VIRTIO_SCSI_VQ_NUM_FIXED + vs->conf.num_queues;
>> +vsc->dev.vqs = s->vhost_vqs;
>> +vsc->dev.vq_index = 0;
>> +vsc->dev.backend_features = 0;
>> +
>> +ret = vhost_dev_init(>dev, >vhost_user,
>> VHOST_BACKEND_TYPE_USER, 0,
>> + errp);
>> +if (ret < 0) {
>> +return ret;
>> +}
>> +
>> +/* restore vhost state */
>> +if (virtio_device_started(vdev, vdev->status)) {
>> +ret = vhost_user_scsi_start(s);
>> +if (ret < 0) {
>> +return ret;
>> +}
>
> Fails without setting an error, violating the function's (tacit)
> postcondition. Callers:
>
> * vhost_user_scsi_event()
>
> Dereferences the null error and crashes.
>
> * vhost_user_scsi_realize_connect()
>
> Also fails without setting an error. Caller:
>
> - vhost_user_scsi_realize()
>
>1. Doesn't emit the "Reconnecting after error: " message. See
> also below.
>
>2. Succeeds instead of failing!
Ack.
>
>> +}
>> +
>> +return 0;
>> +}
>>
Re: [PATCH v3 5/5] vhost-user-scsi: start vhost when guest kicks
> On 1 Sep 2023, at 7:44 PM, Markus Armbruster wrote:
>
> Li Feng mailto:fen...@smartx.com>> writes:
>
>> Let's keep the same behavior as vhost-user-blk.
>>
>> Some old guests kick virtqueue before setting VIRTIO_CONFIG_S_DRIVER_OK.
>>
>> Signed-off-by: Li Feng
>> ---
>> hw/scsi/vhost-user-scsi.c | 48 +++
>> 1 file changed, 44 insertions(+), 4 deletions(-)
>>
>> diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
>> index 5bf012461b..a7fa8e8df2 100644
>> --- a/hw/scsi/vhost-user-scsi.c
>> +++ b/hw/scsi/vhost-user-scsi.c
>> @@ -113,8 +113,48 @@ static void vhost_user_scsi_reset(VirtIODevice *vdev)
>> }
>> }
>>
>> -static void vhost_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
>> +static void vhost_user_scsi_handle_output(VirtIODevice *vdev, VirtQueue *vq)
>> {
>> +VHostUserSCSI *s = (VHostUserSCSI *)vdev;
>> +DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
>> +VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
>> +VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
>> +
>> +Error *local_err = NULL;
>> +int i, ret;
>> +
>> +if (!vdev->start_on_kick) {
>> +return;
>> +}
>> +
>> +if (!s->connected) {
>> +return;
>> +}
>> +
>> +if (vhost_dev_is_started(>dev)) {
>> +return;
>> +}
>> +
>> +/*
>> + * Some guests kick before setting VIRTIO_CONFIG_S_DRIVER_OK so start
>> + * vhost here instead of waiting for .set_status().
>> + */
>> +ret = vhost_user_scsi_start(s);
>> +if (ret < 0) {
>> +error_reportf_err(local_err, "vhost-user-scsi: vhost start failed:
>> ");
>
> Crashes, since @local_err is null. Please test your error paths.
>
> Obvious fix: drop this call.
Emmm, actually I have tested the error path, so I find some issues that I have
fixed
in the following patches.
I will merge the later series into this series.
>
>> +qemu_chr_fe_disconnect(>conf.chardev);
>> +return;
>> +}
>> +
>> +/* Kick right away to begin processing requests already in vring */
>> +for (i = 0; i < vsc->dev.nvqs; i++) {
>> +VirtQueue *kick_vq = virtio_get_queue(vdev, i);
>> +
>> +if (!virtio_queue_get_desc_addr(vdev, i)) {
>> +continue;
>> +}
>> +event_notifier_set(virtio_queue_get_host_notifier(kick_vq));
>> +}
>> }
>>
>> static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
>> @@ -243,9 +283,9 @@ static void vhost_user_scsi_realize(DeviceState *dev,
>> Error **errp)
>> return;
>> }
>>
>> -virtio_scsi_common_realize(dev, vhost_dummy_handle_output,
>> - vhost_dummy_handle_output,
>> - vhost_dummy_handle_output, );
>> +virtio_scsi_common_realize(dev, vhost_user_scsi_handle_output,
>> + vhost_user_scsi_handle_output,
>> + vhost_user_scsi_handle_output, );
>> if (err != NULL) {
>> error_propagate(errp, err);
>> return;
[PATCH v3 2/2] vhost: Add Error parameter to vhost_scsi_common_start()
Add a Error parameter to report the real error, like vhost-user-blk.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 16 +---
hw/scsi/vhost-scsi.c | 5 +++--
hw/scsi/vhost-user-scsi.c | 14 --
include/hw/virtio/vhost-scsi-common.h | 2 +-
4 files changed, 21 insertions(+), 16 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a61cd0e907..4c8637045d 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -16,6 +16,7 @@
*/
#include "qemu/osdep.h"
+#include "qapi/error.h"
#include "qemu/error-report.h"
#include "qemu/module.h"
#include "hw/virtio/vhost.h"
@@ -25,7 +26,7 @@
#include "hw/virtio/virtio-access.h"
#include "hw/fw-path-provider.h"
-int vhost_scsi_common_start(VHostSCSICommon *vsc)
+int vhost_scsi_common_start(VHostSCSICommon *vsc, Error **errp)
{
int ret, i;
VirtIODevice *vdev = VIRTIO_DEVICE(vsc);
@@ -35,18 +36,19 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
VirtIOSCSICommon *vs = (VirtIOSCSICommon *)vsc;
if (!k->set_guest_notifiers) {
-error_report("binding does not support guest notifiers");
+error_setg(errp, "binding does not support guest notifiers");
return -ENOSYS;
}
ret = vhost_dev_enable_notifiers(>dev, vdev);
if (ret < 0) {
+error_setg_errno(errp, -ret, "Error enabling host notifiers");
return ret;
}
ret = k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, true);
if (ret < 0) {
-error_report("Error binding guest notifier");
+error_setg_errno(errp, -ret, "Error binding guest notifier");
goto err_host_notifiers;
}
@@ -54,7 +56,7 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error setting inflight format: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight format");
goto err_guest_notifiers;
}
@@ -64,21 +66,21 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vs->conf.virtqueue_size,
vsc->inflight);
if (ret < 0) {
-error_report("Error getting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error getting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_set_inflight(>dev, vsc->inflight);
if (ret < 0) {
-error_report("Error setting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_start(>dev, vdev, true);
if (ret < 0) {
-error_report("Error start vhost dev");
+error_setg_errno(errp, -ret, "Error starting vhost dev");
goto err_guest_notifiers;
}
diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c
index 443f67daa4..01a3ab4277 100644
--- a/hw/scsi/vhost-scsi.c
+++ b/hw/scsi/vhost-scsi.c
@@ -75,6 +75,7 @@ static int vhost_scsi_start(VHostSCSI *s)
int ret, abi_version;
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
const VhostOps *vhost_ops = vsc->dev.vhost_ops;
+Error *local_err = NULL;
ret = vhost_ops->vhost_scsi_get_abi_version(>dev, _version);
if (ret < 0) {
@@ -88,14 +89,14 @@ static int vhost_scsi_start(VHostSCSI *s)
return -ENOSYS;
}
-ret = vhost_scsi_common_start(vsc);
+ret = vhost_scsi_common_start(vsc, _err);
if (ret < 0) {
return ret;
}
ret = vhost_scsi_set_endpoint(s);
if (ret < 0) {
-error_report("Error setting vhost-scsi endpoint");
+error_reportf_err(local_err, "Error setting vhost-scsi endpoint");
vhost_scsi_common_stop(vsc);
}
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index e931df9f5b..62fc98bb1c 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -43,12 +43,12 @@ enum VhostUserProtocolFeature {
VHOST_USER_PROTOCOL_F_RESET_DEVICE = 13,
};
-static int vhost_user_scsi_start(VHostUserSCSI *s)
+static int vhost_user_scsi_start(VHostUserSCSI *s, Error **errp)
{
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
int ret;
-ret = vhost_scsi_common_start(vsc);
+ret = vhost_scsi_common_start(vsc, errp);
s->started_vu = (ret < 0 ? false : true);
return ret;
@@ -73,6 +73,7 @@ static void vhost_user_scsi_set_status(VirtIODevice *vdev,
uint8_t status)
VHostSCSICommon *vsc = VHOST_SCSI_C
[PATCH v3 1/2] vhost-user: Fix lost reconnect
When the vhost-user is reconnecting to the backend, and if the vhost-user fails
at the get_features in vhost_dev_init(), then the reconnect will fail
and it will not be retriggered forever.
The reason is:
When the vhost-user fails at get_features, the vhost_dev_cleanup will be called
immediately.
vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
The reconnect path is:
vhost_user_blk_event
vhost_user_async_close(.. vhost_user_blk_disconnect ..)
qemu_chr_fe_set_handlers <- clear the notifier callback
schedule vhost_user_async_close_bh
The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
called, then the event fd callback will not be reinstalled.
All vhost-user devices have this issue, including vhost-user-blk/scsi.
With this patch, if the vdev->vdev is null, the fd callback will still
be reinstalled.
Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 2 +-
hw/scsi/vhost-user-scsi.c | 3 ++-
hw/virtio/vhost-user-gpio.c| 2 +-
hw/virtio/vhost-user.c | 9 +++--
include/hw/virtio/vhost-user.h | 3 ++-
5 files changed, 13 insertions(+), 6 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 3c69fa47d5..95c758200d 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -391,7 +391,7 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent
event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >dev,
- vhost_user_blk_disconnect);
+ vhost_user_blk_disconnect,
vhost_user_blk_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index a7fa8e8df2..e931df9f5b 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -236,7 +236,8 @@ static void vhost_user_scsi_event(void *opaque,
QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >conf.chardev, >dev,
- vhost_user_scsi_disconnect);
+ vhost_user_scsi_disconnect,
+ vhost_user_scsi_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index d9979aa5db..04c2cc79f4 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -283,7 +283,7 @@ static void vu_gpio_event(void *opaque, QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >vhost_dev,
- vu_gpio_disconnect);
+ vu_gpio_disconnect, vu_gpio_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index 8dcf049d42..12c4a41f30 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2643,6 +2643,7 @@ typedef struct {
DeviceState *dev;
CharBackend *cd;
struct vhost_dev *vhost;
+IOEventHandler *event_cb;
} VhostAsyncCallback;
static void vhost_user_async_close_bh(void *opaque)
@@ -2657,7 +2658,10 @@ static void vhost_user_async_close_bh(void *opaque)
*/
if (vhost->vdev) {
data->cb(data->dev);
-}
+} else if (data->event_cb) {
+qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
+ NULL, data->dev, NULL, true);
+ }
g_free(data);
}
@@ -2669,7 +2673,8 @@ static void vhost_user_async_close_bh(void *opaque)
*/
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb)
+vu_async_close_fn cb,
+IOEventHandler *event_cb)
{
if (!runstate_check(RUN_STATE_SHUTDOWN)) {
/*
diff --git a/include/hw/virtio/vhost-user.h b/include/hw/virtio/vhost-user.h
index 191216a74f..649e9dd54f 100644
--- a/include/hw/virtio/vhost-user.h
+++ b/include/hw/virtio/vhost-user.h
@@ -84,6 +84,7 @@ typedef void (*vu_async_close_fn)(DeviceState *cb);
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb);
+vu_async_close_fn cb,
+IOEventHandler *event_cb);
#endif
--
2.41.0
[PATCH v3 0/2] Fix vhost reconnect issues
The patchset fixes the regression issue of vhost reconnect. It's a serious bug that the vhost-user will lose the reconnect forever. The 2nd patch enhances the error handle of vhost-user-scsi. This patchset's parent commit is: https://lore.kernel.org/all/20230731121018.2856310-1-fen...@smartx.com/ Changes for v3: - Fix the code style. Changes for v2: - Add a event_cb in VhostAsyncCallback to be called when dev is NULL; - Fix the error report message. Li Feng (2): vhost-user: Fix lost reconnect vhost: Add Error parameter to vhost_scsi_common_start() hw/block/vhost-user-blk.c | 2 +- hw/scsi/vhost-scsi-common.c | 16 +--- hw/scsi/vhost-scsi.c | 5 +++-- hw/scsi/vhost-user-scsi.c | 17 ++--- hw/virtio/vhost-user-gpio.c | 2 +- hw/virtio/vhost-user.c| 9 +++-- include/hw/virtio/vhost-scsi-common.h | 2 +- include/hw/virtio/vhost-user.h| 3 ++- 8 files changed, 34 insertions(+), 22 deletions(-) -- 2.41.0
Re: [PATCH v2 1/2] vhost-user: Fix lost reconnect
> On 30 Aug 2023, at 6:11 AM, Raphael Norwitz
> wrote:
>
>
>
>> On Aug 24, 2023, at 3:41 AM, Li Feng wrote:
>>
>> When the vhost-user is reconnecting to the backend, and if the vhost-user
>> fails
>> at the get_features in vhost_dev_init(), then the reconnect will fail
>> and it will not be retriggered forever.
>>
>> The reason is:
>> When the vhost-user fails at get_features, the vhost_dev_cleanup will be
>> called
>> immediately.
>>
>> vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
>>
>> The reconnect path is:
>> vhost_user_blk_event
>> vhost_user_async_close(.. vhost_user_blk_disconnect ..)
>>qemu_chr_fe_set_handlers <- clear the notifier callback
>> schedule vhost_user_async_close_bh
>>
>> The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
>> called, then the event fd callback will not be reinstalled.
>>
>> All vhost-user devices have this issue, including vhost-user-blk/scsi.
>>
>> With this patch, if the vdev->vdev is null, the fd callback will still
>> be reinstalled.
>>
>> Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
>>
>
> A couple of NITs, otherwise LGTM
>
> Reviewed-by: Raphael Norwitz <mailto:raphael.norw...@nutanix.com>>
>
>> Signed-off-by: Li Feng
>> ---
>> hw/block/vhost-user-blk.c | 2 +-
>> hw/scsi/vhost-user-scsi.c | 3 ++-
>> hw/virtio/vhost-user-gpio.c| 2 +-
>> hw/virtio/vhost-user.c | 10 --
>> include/hw/virtio/vhost-user.h | 4 +++-
>> 5 files changed, 15 insertions(+), 6 deletions(-)
>>
>> diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
>> index 3c69fa47d5..95c758200d 100644
>> --- a/hw/block/vhost-user-blk.c
>> +++ b/hw/block/vhost-user-blk.c
>> @@ -391,7 +391,7 @@ static void vhost_user_blk_event(void *opaque,
>> QEMUChrEvent event)
>>case CHR_EVENT_CLOSED:
>>/* defer close until later to avoid circular close */
>>vhost_user_async_close(dev, >chardev, >dev,
>> - vhost_user_blk_disconnect);
>> + vhost_user_blk_disconnect,
>> vhost_user_blk_event);
>>break;
>>case CHR_EVENT_BREAK:
>>case CHR_EVENT_MUX_IN:
>> diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
>> index a7fa8e8df2..e931df9f5b 100644
>> --- a/hw/scsi/vhost-user-scsi.c
>> +++ b/hw/scsi/vhost-user-scsi.c
>> @@ -236,7 +236,8 @@ static void vhost_user_scsi_event(void *opaque,
>> QEMUChrEvent event)
>>case CHR_EVENT_CLOSED:
>>/* defer close until later to avoid circular close */
>>vhost_user_async_close(dev, >conf.chardev, >dev,
>> - vhost_user_scsi_disconnect);
>> + vhost_user_scsi_disconnect,
>> + vhost_user_scsi_event);
>>break;
>>case CHR_EVENT_BREAK:
>>case CHR_EVENT_MUX_IN:
>> diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
>> index d9979aa5db..04c2cc79f4 100644
>> --- a/hw/virtio/vhost-user-gpio.c
>> +++ b/hw/virtio/vhost-user-gpio.c
>> @@ -283,7 +283,7 @@ static void vu_gpio_event(void *opaque, QEMUChrEvent
>> event)
>>case CHR_EVENT_CLOSED:
>>/* defer close until later to avoid circular close */
>>vhost_user_async_close(dev, >chardev, >vhost_dev,
>> - vu_gpio_disconnect);
>> + vu_gpio_disconnect, vu_gpio_event);
>>break;
>>case CHR_EVENT_BREAK:
>>case CHR_EVENT_MUX_IN:
>> diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
>> index 8dcf049d42..9540766dd3 100644
>> --- a/hw/virtio/vhost-user.c
>> +++ b/hw/virtio/vhost-user.c
>> @@ -2643,6 +2643,7 @@ typedef struct {
>>DeviceState *dev;
>>CharBackend *cd;
>>struct vhost_dev *vhost;
>> +IOEventHandler *event_cb;
>> } VhostAsyncCallback;
>>
>> static void vhost_user_async_close_bh(void *opaque)
>> @@ -2657,7 +2658,10 @@ static void vhost_user_async_close_bh(void *opaque)
>> */
>>if (vhost->vdev) {
>>data->cb(data->dev);
>> -}
>> +} else if (data->event_cb) {
>> +qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
>> + NULL, data->dev, NULL, true);
>> + }
&
[PATCH v2 1/2] vhost-user: Fix lost reconnect
When the vhost-user is reconnecting to the backend, and if the vhost-user fails
at the get_features in vhost_dev_init(), then the reconnect will fail
and it will not be retriggered forever.
The reason is:
When the vhost-user fails at get_features, the vhost_dev_cleanup will be called
immediately.
vhost_dev_cleanup calls 'memset(hdev, 0, sizeof(struct vhost_dev))'.
The reconnect path is:
vhost_user_blk_event
vhost_user_async_close(.. vhost_user_blk_disconnect ..)
qemu_chr_fe_set_handlers <- clear the notifier callback
schedule vhost_user_async_close_bh
The vhost->vdev is null, so the vhost_user_blk_disconnect will not be
called, then the event fd callback will not be reinstalled.
All vhost-user devices have this issue, including vhost-user-blk/scsi.
With this patch, if the vdev->vdev is null, the fd callback will still
be reinstalled.
Fixes: 71e076a07d ("hw/virtio: generalise CHR_EVENT_CLOSED handling")
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 2 +-
hw/scsi/vhost-user-scsi.c | 3 ++-
hw/virtio/vhost-user-gpio.c| 2 +-
hw/virtio/vhost-user.c | 10 --
include/hw/virtio/vhost-user.h | 4 +++-
5 files changed, 15 insertions(+), 6 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 3c69fa47d5..95c758200d 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -391,7 +391,7 @@ static void vhost_user_blk_event(void *opaque, QEMUChrEvent
event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >dev,
- vhost_user_blk_disconnect);
+ vhost_user_blk_disconnect,
vhost_user_blk_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index a7fa8e8df2..e931df9f5b 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -236,7 +236,8 @@ static void vhost_user_scsi_event(void *opaque,
QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >conf.chardev, >dev,
- vhost_user_scsi_disconnect);
+ vhost_user_scsi_disconnect,
+ vhost_user_scsi_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index d9979aa5db..04c2cc79f4 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -283,7 +283,7 @@ static void vu_gpio_event(void *opaque, QEMUChrEvent event)
case CHR_EVENT_CLOSED:
/* defer close until later to avoid circular close */
vhost_user_async_close(dev, >chardev, >vhost_dev,
- vu_gpio_disconnect);
+ vu_gpio_disconnect, vu_gpio_event);
break;
case CHR_EVENT_BREAK:
case CHR_EVENT_MUX_IN:
diff --git a/hw/virtio/vhost-user.c b/hw/virtio/vhost-user.c
index 8dcf049d42..9540766dd3 100644
--- a/hw/virtio/vhost-user.c
+++ b/hw/virtio/vhost-user.c
@@ -2643,6 +2643,7 @@ typedef struct {
DeviceState *dev;
CharBackend *cd;
struct vhost_dev *vhost;
+IOEventHandler *event_cb;
} VhostAsyncCallback;
static void vhost_user_async_close_bh(void *opaque)
@@ -2657,7 +2658,10 @@ static void vhost_user_async_close_bh(void *opaque)
*/
if (vhost->vdev) {
data->cb(data->dev);
-}
+} else if (data->event_cb) {
+qemu_chr_fe_set_handlers(data->cd, NULL, NULL, data->event_cb,
+ NULL, data->dev, NULL, true);
+ }
g_free(data);
}
@@ -2669,7 +2673,9 @@ static void vhost_user_async_close_bh(void *opaque)
*/
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb)
+vu_async_close_fn cb,
+IOEventHandler *event_cb
+)
{
if (!runstate_check(RUN_STATE_SHUTDOWN)) {
/*
diff --git a/include/hw/virtio/vhost-user.h b/include/hw/virtio/vhost-user.h
index 191216a74f..5fdc711d4e 100644
--- a/include/hw/virtio/vhost-user.h
+++ b/include/hw/virtio/vhost-user.h
@@ -84,6 +84,8 @@ typedef void (*vu_async_close_fn)(DeviceState *cb);
void vhost_user_async_close(DeviceState *d,
CharBackend *chardev, struct vhost_dev *vhost,
-vu_async_close_fn cb);
+vu_async_close_fn cb,
+IOEventHandler *event_cb
+);
#endif
--
2.41.0
[PATCH v2 0/2] Fix vhost reconnect issues
The patchset fixes the regression issue of vhost reconnect. It's a serious bug that the vhost-user will lose the reconnect forever. The 2nd patch enhances the error handle of vhost-user-scsi. This patchset's parent commit is: https://lore.kernel.org/all/20230731121018.2856310-1-fen...@smartx.com/ Changes for v2: - Add a event_cb in VhostAsyncCallback to be called when dev is NULL; - Fix the error report message. Li Feng (2): vhost-user: Fix lost reconnect vhost: Add Error parameter to vhost_scsi_common_start() hw/block/vhost-user-blk.c | 2 +- hw/scsi/vhost-scsi-common.c | 16 +--- hw/scsi/vhost-scsi.c | 5 +++-- hw/scsi/vhost-user-scsi.c | 17 ++--- hw/virtio/vhost-user-gpio.c | 2 +- hw/virtio/vhost-user.c| 10 -- include/hw/virtio/vhost-scsi-common.h | 2 +- include/hw/virtio/vhost-user.h| 4 +++- 8 files changed, 36 insertions(+), 22 deletions(-) -- 2.41.0
[PATCH v2 2/2] vhost: Add Error parameter to vhost_scsi_common_start()
Add a Error parameter to report the real error, like vhost-user-blk.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 16 +---
hw/scsi/vhost-scsi.c | 5 +++--
hw/scsi/vhost-user-scsi.c | 14 --
include/hw/virtio/vhost-scsi-common.h | 2 +-
4 files changed, 21 insertions(+), 16 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a61cd0e907..4c8637045d 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -16,6 +16,7 @@
*/
#include "qemu/osdep.h"
+#include "qapi/error.h"
#include "qemu/error-report.h"
#include "qemu/module.h"
#include "hw/virtio/vhost.h"
@@ -25,7 +26,7 @@
#include "hw/virtio/virtio-access.h"
#include "hw/fw-path-provider.h"
-int vhost_scsi_common_start(VHostSCSICommon *vsc)
+int vhost_scsi_common_start(VHostSCSICommon *vsc, Error **errp)
{
int ret, i;
VirtIODevice *vdev = VIRTIO_DEVICE(vsc);
@@ -35,18 +36,19 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
VirtIOSCSICommon *vs = (VirtIOSCSICommon *)vsc;
if (!k->set_guest_notifiers) {
-error_report("binding does not support guest notifiers");
+error_setg(errp, "binding does not support guest notifiers");
return -ENOSYS;
}
ret = vhost_dev_enable_notifiers(>dev, vdev);
if (ret < 0) {
+error_setg_errno(errp, -ret, "Error enabling host notifiers");
return ret;
}
ret = k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, true);
if (ret < 0) {
-error_report("Error binding guest notifier");
+error_setg_errno(errp, -ret, "Error binding guest notifier");
goto err_host_notifiers;
}
@@ -54,7 +56,7 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error setting inflight format: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight format");
goto err_guest_notifiers;
}
@@ -64,21 +66,21 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vs->conf.virtqueue_size,
vsc->inflight);
if (ret < 0) {
-error_report("Error getting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error getting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_set_inflight(>dev, vsc->inflight);
if (ret < 0) {
-error_report("Error setting inflight: %d", -ret);
+error_setg_errno(errp, -ret, "Error setting inflight");
goto err_guest_notifiers;
}
}
ret = vhost_dev_start(>dev, vdev, true);
if (ret < 0) {
-error_report("Error start vhost dev");
+error_setg_errno(errp, -ret, "Error starting vhost dev");
goto err_guest_notifiers;
}
diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c
index 443f67daa4..01a3ab4277 100644
--- a/hw/scsi/vhost-scsi.c
+++ b/hw/scsi/vhost-scsi.c
@@ -75,6 +75,7 @@ static int vhost_scsi_start(VHostSCSI *s)
int ret, abi_version;
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
const VhostOps *vhost_ops = vsc->dev.vhost_ops;
+Error *local_err = NULL;
ret = vhost_ops->vhost_scsi_get_abi_version(>dev, _version);
if (ret < 0) {
@@ -88,14 +89,14 @@ static int vhost_scsi_start(VHostSCSI *s)
return -ENOSYS;
}
-ret = vhost_scsi_common_start(vsc);
+ret = vhost_scsi_common_start(vsc, _err);
if (ret < 0) {
return ret;
}
ret = vhost_scsi_set_endpoint(s);
if (ret < 0) {
-error_report("Error setting vhost-scsi endpoint");
+error_reportf_err(local_err, "Error setting vhost-scsi endpoint");
vhost_scsi_common_stop(vsc);
}
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index e931df9f5b..62fc98bb1c 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -43,12 +43,12 @@ enum VhostUserProtocolFeature {
VHOST_USER_PROTOCOL_F_RESET_DEVICE = 13,
};
-static int vhost_user_scsi_start(VHostUserSCSI *s)
+static int vhost_user_scsi_start(VHostUserSCSI *s, Error **errp)
{
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
int ret;
-ret = vhost_scsi_common_start(vsc);
+ret = vhost_scsi_common_start(vsc, errp);
s->started_vu = (ret < 0 ? false : true);
return ret;
@@ -73,6 +73,7 @@ static void vhost_user_scsi_set_status(VirtIODevice *vdev,
uint8_t status)
VHostSCSICommon *vsc = VHOST_SCSI_C
[PATCH v3 0/5] Implement reconnect for vhost-user-scsi
This patchset adds reconnect support for vhost-user-scsi. At the same times, fix vhost fd leak and refactor some code. Changes for v3: - Split the vhost_user_scsi_handle_output to a separate patch; - Move the started_vu from vhost scsi common header to vhost-user-scsi header; - Fix a log print error; Changes for v2: - Split the v1 patch to small separate patchset; - New patch for fixing fd leak, which has sent to reviewers in another mail; - Implement the `vhost_user_scsi_handle_output`; - Add the started_vu safe check; - Fix error handler; - Check the inflight before set/get inflight fd. Li Feng (5): vhost: fix the fd leak vhost-user-common: send get_inflight_fd once vhost: move and rename the conn retry times vhost-user-scsi: support reconnect to backend vhost-user-scsi: start vhost when guest kicks hw/block/vhost-user-blk.c | 4 +- hw/scsi/vhost-scsi-common.c | 37 ++--- hw/scsi/vhost-user-scsi.c | 247 +--- hw/virtio/vhost-user-gpio.c | 3 +- hw/virtio/vhost.c | 2 + include/hw/virtio/vhost-user-scsi.h | 4 + include/hw/virtio/vhost.h | 2 + 7 files changed, 252 insertions(+), 47 deletions(-) -- 2.41.0
[PATCH v3 5/5] vhost-user-scsi: start vhost when guest kicks
Let's keep the same behavior as vhost-user-blk.
Some old guests kick virtqueue before setting VIRTIO_CONFIG_S_DRIVER_OK.
Signed-off-by: Li Feng
---
hw/scsi/vhost-user-scsi.c | 48 +++
1 file changed, 44 insertions(+), 4 deletions(-)
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 5bf012461b..a7fa8e8df2 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -113,8 +113,48 @@ static void vhost_user_scsi_reset(VirtIODevice *vdev)
}
}
-static void vhost_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+static void vhost_user_scsi_handle_output(VirtIODevice *vdev, VirtQueue *vq)
{
+VHostUserSCSI *s = (VHostUserSCSI *)vdev;
+DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+
+Error *local_err = NULL;
+int i, ret;
+
+if (!vdev->start_on_kick) {
+return;
+}
+
+if (!s->connected) {
+return;
+}
+
+if (vhost_dev_is_started(>dev)) {
+return;
+}
+
+/*
+ * Some guests kick before setting VIRTIO_CONFIG_S_DRIVER_OK so start
+ * vhost here instead of waiting for .set_status().
+ */
+ret = vhost_user_scsi_start(s);
+if (ret < 0) {
+error_reportf_err(local_err, "vhost-user-scsi: vhost start failed: ");
+qemu_chr_fe_disconnect(>conf.chardev);
+return;
+}
+
+/* Kick right away to begin processing requests already in vring */
+for (i = 0; i < vsc->dev.nvqs; i++) {
+VirtQueue *kick_vq = virtio_get_queue(vdev, i);
+
+if (!virtio_queue_get_desc_addr(vdev, i)) {
+continue;
+}
+event_notifier_set(virtio_queue_get_host_notifier(kick_vq));
+}
}
static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
@@ -243,9 +283,9 @@ static void vhost_user_scsi_realize(DeviceState *dev, Error
**errp)
return;
}
-virtio_scsi_common_realize(dev, vhost_dummy_handle_output,
- vhost_dummy_handle_output,
- vhost_dummy_handle_output, );
+virtio_scsi_common_realize(dev, vhost_user_scsi_handle_output,
+ vhost_user_scsi_handle_output,
+ vhost_user_scsi_handle_output, );
if (err != NULL) {
error_propagate(errp, err);
return;
--
2.41.0
[PATCH v3 1/5] vhost: fix the fd leak
When the vhost-user reconnect to the backend, the notifer should be
cleanup. Otherwise, the fd resource will be exhausted.
Fixes: f9a09ca3ea ("vhost: add support for configure interrupt")
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/virtio/vhost.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c
index abf0d03c8d..e2f6ffb446 100644
--- a/hw/virtio/vhost.c
+++ b/hw/virtio/vhost.c
@@ -2044,6 +2044,8 @@ void vhost_dev_stop(struct vhost_dev *hdev, VirtIODevice
*vdev, bool vrings)
event_notifier_test_and_clear(
>vqs[VHOST_QUEUE_NUM_CONFIG_INR].masked_config_notifier);
event_notifier_test_and_clear(>config_notifier);
+event_notifier_cleanup(
+>vqs[VHOST_QUEUE_NUM_CONFIG_INR].masked_config_notifier);
trace_vhost_dev_stop(hdev, vdev->name, vrings);
--
2.41.0
[PATCH v3 4/5] vhost-user-scsi: support reconnect to backend
If the backend crashes and restarts, the device is broken.
This patch adds reconnect for vhost-user-scsi.
Tested with spdk backend.
Signed-off-by: Li Feng
---
hw/scsi/vhost-user-scsi.c | 199 +---
include/hw/virtio/vhost-user-scsi.h | 4 +
2 files changed, 184 insertions(+), 19 deletions(-)
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index ee99b19e7a..5bf012461b 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -43,26 +43,54 @@ enum VhostUserProtocolFeature {
VHOST_USER_PROTOCOL_F_RESET_DEVICE = 13,
};
+static int vhost_user_scsi_start(VHostUserSCSI *s)
+{
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+int ret;
+
+ret = vhost_scsi_common_start(vsc);
+s->started_vu = (ret < 0 ? false : true);
+
+return ret;
+}
+
+static void vhost_user_scsi_stop(VHostUserSCSI *s)
+{
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+
+if (!s->started_vu) {
+return;
+}
+s->started_vu = false;
+
+vhost_scsi_common_stop(vsc);
+}
+
static void vhost_user_scsi_set_status(VirtIODevice *vdev, uint8_t status)
{
VHostUserSCSI *s = (VHostUserSCSI *)vdev;
+DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
-bool start = (status & VIRTIO_CONFIG_S_DRIVER_OK) && vdev->vm_running;
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+bool should_start = virtio_device_should_start(vdev, status);
+int ret;
-if (vhost_dev_is_started(>dev) == start) {
+if (!s->connected) {
return;
}
-if (start) {
-int ret;
+if (vhost_dev_is_started(>dev) == should_start) {
+return;
+}
-ret = vhost_scsi_common_start(vsc);
+if (should_start) {
+ret = vhost_user_scsi_start(s);
if (ret < 0) {
error_report("unable to start vhost-user-scsi: %s",
strerror(-ret));
-exit(1);
+qemu_chr_fe_disconnect(>conf.chardev);
}
} else {
-vhost_scsi_common_stop(vsc);
+vhost_user_scsi_stop(s);
}
}
@@ -89,14 +117,126 @@ static void vhost_dummy_handle_output(VirtIODevice *vdev,
VirtQueue *vq)
{
}
+static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
+{
+VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+VHostUserSCSI *s = VHOST_USER_SCSI(vdev);
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+int ret = 0;
+
+if (s->connected) {
+return 0;
+}
+s->connected = true;
+
+vsc->dev.num_queues = vs->conf.num_queues;
+vsc->dev.nvqs = VIRTIO_SCSI_VQ_NUM_FIXED + vs->conf.num_queues;
+vsc->dev.vqs = s->vhost_vqs;
+vsc->dev.vq_index = 0;
+vsc->dev.backend_features = 0;
+
+ret = vhost_dev_init(>dev, >vhost_user, VHOST_BACKEND_TYPE_USER, 0,
+ errp);
+if (ret < 0) {
+return ret;
+}
+
+/* restore vhost state */
+if (virtio_device_started(vdev, vdev->status)) {
+ret = vhost_user_scsi_start(s);
+if (ret < 0) {
+return ret;
+}
+}
+
+return 0;
+}
+
+static void vhost_user_scsi_event(void *opaque, QEMUChrEvent event);
+
+static void vhost_user_scsi_disconnect(DeviceState *dev)
+{
+VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+VHostUserSCSI *s = VHOST_USER_SCSI(vdev);
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+
+if (!s->connected) {
+return;
+}
+s->connected = false;
+
+vhost_user_scsi_stop(s);
+
+vhost_dev_cleanup(>dev);
+
+/* Re-instate the event handler for new connections */
+qemu_chr_fe_set_handlers(>conf.chardev, NULL, NULL,
+ vhost_user_scsi_event, NULL, dev, NULL, true);
+}
+
+static void vhost_user_scsi_event(void *opaque, QEMUChrEvent event)
+{
+DeviceState *dev = opaque;
+VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+VHostUserSCSI *s = VHOST_USER_SCSI(vdev);
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+Error *local_err = NULL;
+
+switch (event) {
+case CHR_EVENT_OPENED:
+if (vhost_user_scsi_connect(dev, _err) < 0) {
+error_report_err(local_err);
+qemu_chr_fe_disconnect(>conf.chardev);
+return;
+}
+break;
+case CHR_EVENT_CLOSED:
+/* defer close until later to avoid circular close */
+vhost_user_async_close(dev, >conf.chardev, >dev,
+ vhost_user_scsi_disconnect);
+break;
+case CHR_EVENT_BREAK:
+case CHR_EVENT_MUX_IN:
+case CHR_EVENT_MUX_OUT:
+/* Ignore */
+break;
+}
+}
+
+static int vhost_user_scsi_realize_connect(VHostU
[PATCH v3 2/5] vhost-user-common: send get_inflight_fd once
Currently the get_inflight_fd will be sent every time the device is started, and
the backend will allocate shared memory to save the inflight state. If the
backend finds that it receives the second get_inflight_fd, it will release the
previous shared memory, which breaks inflight working logic.
This patch is a preparation for the following patches.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 37 ++---
1 file changed, 18 insertions(+), 19 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a06f01af26..a61cd0e907 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -52,20 +52,28 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vsc->dev.acked_features = vdev->guest_features;
-assert(vsc->inflight == NULL);
-vsc->inflight = g_new0(struct vhost_inflight, 1);
-ret = vhost_dev_get_inflight(>dev,
- vs->conf.virtqueue_size,
- vsc->inflight);
+ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error get inflight: %d", -ret);
+error_report("Error setting inflight format: %d", -ret);
goto err_guest_notifiers;
}
-ret = vhost_dev_set_inflight(>dev, vsc->inflight);
-if (ret < 0) {
-error_report("Error set inflight: %d", -ret);
-goto err_guest_notifiers;
+if (vsc->inflight) {
+if (!vsc->inflight->addr) {
+ret = vhost_dev_get_inflight(>dev,
+vs->conf.virtqueue_size,
+vsc->inflight);
+if (ret < 0) {
+error_report("Error getting inflight: %d", -ret);
+goto err_guest_notifiers;
+}
+}
+
+ret = vhost_dev_set_inflight(>dev, vsc->inflight);
+if (ret < 0) {
+error_report("Error setting inflight: %d", -ret);
+goto err_guest_notifiers;
+}
}
ret = vhost_dev_start(>dev, vdev, true);
@@ -85,9 +93,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
return ret;
err_guest_notifiers:
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-
k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
err_host_notifiers:
vhost_dev_disable_notifiers(>dev, vdev);
@@ -111,12 +116,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
}
assert(ret >= 0);
-if (vsc->inflight) {
-vhost_dev_free_inflight(vsc->inflight);
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-}
-
vhost_dev_disable_notifiers(>dev, vdev);
}
--
2.41.0
[PATCH v3 3/5] vhost: move and rename the conn retry times
Multiple devices need this macro, move it to a common header.
Signed-off-by: Li Feng
Reviewed-by: Raphael Norwitz
---
hw/block/vhost-user-blk.c | 4 +---
hw/virtio/vhost-user-gpio.c | 3 +--
include/hw/virtio/vhost.h | 2 ++
3 files changed, 4 insertions(+), 5 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index eecf3f7a81..3c69fa47d5 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -32,8 +32,6 @@
#include "sysemu/sysemu.h"
#include "sysemu/runstate.h"
-#define REALIZE_CONNECTION_RETRIES 3
-
static const int user_feature_bits[] = {
VIRTIO_BLK_F_SIZE_MAX,
VIRTIO_BLK_F_SEG_MAX,
@@ -482,7 +480,7 @@ static void vhost_user_blk_device_realize(DeviceState *dev,
Error **errp)
s->inflight = g_new0(struct vhost_inflight, 1);
s->vhost_vqs = g_new0(struct vhost_virtqueue, s->num_queues);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
assert(!*errp);
do {
if (*errp) {
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index 3b013f2d0f..d9979aa5db 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -15,7 +15,6 @@
#include "standard-headers/linux/virtio_ids.h"
#include "trace.h"
-#define REALIZE_CONNECTION_RETRIES 3
#define VHOST_NVQS 2
/* Features required from VirtIO */
@@ -359,7 +358,7 @@ static void vu_gpio_device_realize(DeviceState *dev, Error
**errp)
qemu_chr_fe_set_handlers(>chardev, NULL, NULL, vu_gpio_event, NULL,
dev, NULL, true);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
g_assert(!*errp);
do {
if (*errp) {
diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h
index 6a173cb9fa..ca3131b1af 100644
--- a/include/hw/virtio/vhost.h
+++ b/include/hw/virtio/vhost.h
@@ -8,6 +8,8 @@
#define VHOST_F_DEVICE_IOTLB 63
#define VHOST_USER_F_PROTOCOL_FEATURES 30
+#define VU_REALIZE_CONN_RETRIES 3
+
/* Generic structures common for any vhost based device. */
struct vhost_inflight {
--
2.41.0
Re: [PATCH v2 2/4] vhost-user-common: send get_inflight_fd once
> 2023年7月31日 06:13，Raphael Norwitz 写道：
>
>>
>> On Jul 28, 2023, at 3:49 AM, Li Feng wrote:
>>
>>
>>
>>> 2023年7月28日 下午2:04，Michael S. Tsirkin 写道：
>>>
>>> On Tue, Jul 25, 2023 at 06:42:45PM +0800, Li Feng wrote:
>>>> Get_inflight_fd is sent only once. When reconnecting to the backend,
>>>> qemu sent set_inflight_fd to the backend.
>>>
>>> I don't understand what you are trying to say here.
>>> Should be:
>>> Currently ABCD. This is wrong/unnecessary because EFG. This patch HIJ.
>>
>> Thanks, I will reorganize the commit message in v3.
>>>
>>>> Signed-off-by: Li Feng
>>>> ---
>>>> hw/scsi/vhost-scsi-common.c | 37 ++---
>>>> 1 file changed, 18 insertions(+), 19 deletions(-)
>>>>
>>>> diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
>>>> index a06f01af26..664adb15b4 100644
>>>> --- a/hw/scsi/vhost-scsi-common.c
>>>> +++ b/hw/scsi/vhost-scsi-common.c
>>>> @@ -52,20 +52,28 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>>>>
>>>>vsc->dev.acked_features = vdev->guest_features;
>>>>
>>>> -assert(vsc->inflight == NULL);
>>>> -vsc->inflight = g_new0(struct vhost_inflight, 1);
>>>> -ret = vhost_dev_get_inflight(>dev,
>>>> - vs->conf.virtqueue_size,
>>>> - vsc->inflight);
>>>> +ret = vhost_dev_prepare_inflight(>dev, vdev);
>>>>if (ret < 0) {
>>>> -error_report("Error get inflight: %d", -ret);
>>>> +error_report("Error setting inflight format: %d", -ret);
>>>>goto err_guest_notifiers;
>>>>}
>>>>
>>>> -ret = vhost_dev_set_inflight(>dev, vsc->inflight);
>>>> -if (ret < 0) {
>>>> -error_report("Error set inflight: %d", -ret);
>>>> -goto err_guest_notifiers;
>>>> +if (vsc->inflight) {
>>>> +if (!vsc->inflight->addr) {
>>>> +ret = vhost_dev_get_inflight(>dev,
>>>> +vs->conf.virtqueue_size,
>>>> +vsc->inflight);
>>>> +if (ret < 0) {
>>>> +error_report("Error get inflight: %d", -ret);
>>>
>>> As long as you are fixing this - should be "getting inflight”.
>> I will fix it in v3.
>>>
>>>> +goto err_guest_notifiers;
>>>> +}
>>>> +}
>>>> +
>
> Looks like you reworked this a bit so to avoid a potential crash if
> vsc->inflight is NULL
>
> Should we fix it for vhost-user-blk too?
>
This check is mainly for the vhost-scsi code, that doesn’t need allocate the
inflight memory.
The vhost-user-blk doesn’t need this check, because there isn't a vhost-blk
device that reuse the code.
>>>> +ret = vhost_dev_set_inflight(>dev, vsc->inflight);
>>>> +if (ret < 0) {
>>>> +error_report("Error set inflight: %d", -ret);
>>>> +goto err_guest_notifiers;
>>>> +}
>>>>}
>>>>
>>>>ret = vhost_dev_start(>dev, vdev, true);
>>>> @@ -85,9 +93,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>>>>return ret;
>>>>
>>>> err_guest_notifiers:
>>>> -g_free(vsc->inflight);
>>>> -vsc->inflight = NULL;
>>>> -
>>>>k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
>>>> err_host_notifiers:
>>>>vhost_dev_disable_notifiers(>dev, vdev);
>>>> @@ -111,12 +116,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
>>>>}
>>>>assert(ret >= 0);
>>>>
>
> As I said before, I think this introduces a leak.
I have answered in the previous mail.
>
>>>> -if (vsc->inflight) {
>>>> -vhost_dev_free_inflight(vsc->inflight);
>>>> -g_free(vsc->inflight);
>>>> -vsc->inflight = NULL;
>>>> -}
>>>> -
>>>>vhost_dev_disable_notifiers(>dev, vdev);
>>>> }
>>>>
>>>> --
>>>> 2.41.0
Re: [PATCH] vhost-user-scsi: support reconnect to backend
> 2023年7月31日 06:09，Raphael Norwitz 写道：
>
>
>
>> On Jul 28, 2023, at 3:48 AM, Li Feng wrote:
>>
>> Thanks for your reply.
>>
>>> 2023年7月28日 上午5:21，Raphael Norwitz 写道：
>>>
>>>
>>>
>>>> On Jul 25, 2023, at 6:19 AM, Li Feng wrote:
>>>>
>>>> Thanks for your comments.
>>>>
>>>>> 2023年7月25日 上午1:21，Raphael Norwitz 写道：
>>>>>
>>>>> Very excited to see this. High level looks good modulo a few small things.
>>>>>
>>>>> My major concern is around existing vhost-user-scsi backends which don’t
>>>>> support VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD. IMO we should hide the
>>>>> reconnect behavior behind a VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD check.
>>>>> We may want to do the same for vhost-user-blk.
>>>>>
>>>>> The question is then what happens if the check is false. IIUC without an
>>>>> inflight FD, if a device processes requests out of order, it’s not safe
>>>>> to continue execution on reconnect, as there’s no way for the backend to
>>>>> know how to replay IO. Should we permanently wedge the device or have
>>>>> QEMU fail out? May be nice to have a toggle for this.
>>>>
>>>> Based on what MST said, is there anything else I need to do?
>>>
>>> I don’t think so.
>>>
>>>>>
>>>>>> On Jul 21, 2023, at 6:51 AM, Li Feng wrote:
>>>>>>
>>>>>> If the backend crashes and restarts, the device is broken.
>>>>>> This patch adds reconnect for vhost-user-scsi.
>>>>>>
>>>>>> Tested with spdk backend.
>>>>>>
>>>>>> Signed-off-by: Li Feng
>>>>>> ---
>>>>>> hw/block/vhost-user-blk.c | 2 -
>>>>>> hw/scsi/vhost-scsi-common.c | 27 ++---
>>>>>> hw/scsi/vhost-user-scsi.c | 163 +---
>>>>>> include/hw/virtio/vhost-user-scsi.h | 3 +
>>>>>> include/hw/virtio/vhost.h | 2 +
>>>>>> 5 files changed, 165 insertions(+), 32 deletions(-)
>>>>>>
>>>>>> diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
>>>>>> index eecf3f7a81..f250c740b5 100644
>>>>>> --- a/hw/block/vhost-user-blk.c
>>>>>> +++ b/hw/block/vhost-user-blk.c
>>>>>> @@ -32,8 +32,6 @@
>>>>>> #include "sysemu/sysemu.h"
>>>>>> #include "sysemu/runstate.h"
>>>>>>
>>>>>> -#define REALIZE_CONNECTION_RETRIES 3
>>>>>> -
>>>>>> static const int user_feature_bits[] = {
>>>>>> VIRTIO_BLK_F_SIZE_MAX,
>>>>>> VIRTIO_BLK_F_SEG_MAX,
>>>>>> diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
>>>>>
>>>>> Why can’t all the vhost-scsi-common stuff be moved to a separate change?
>>>>
>>>> I will move this code to separate patch.
>>>>>
>>>>> Especially the stuff introduced for vhost-user-blk in
>>>>> 1b0063b3048af65dfaae6422a572c87db8575a92 should be moved out.
>>>> OK.
>>>>
>>>>>
>>>>>> index a06f01af26..08801886b8 100644
>>>>>> --- a/hw/scsi/vhost-scsi-common.c
>>>>>> +++ b/hw/scsi/vhost-scsi-common.c
>>>>>> @@ -52,16 +52,22 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>>>>>>
>>>>>> vsc->dev.acked_features = vdev->guest_features;
>>>>>>
>>>>>> -assert(vsc->inflight == NULL);
>>>>>> -vsc->inflight = g_new0(struct vhost_inflight, 1);
>>>>>> -ret = vhost_dev_get_inflight(>dev,
>>>>>> - vs->conf.virtqueue_size,
>>>>>> - vsc->inflight);
>>>>>> +ret = vhost_dev_prepare_inflight(>dev, vdev);
>>>>>> if (ret < 0) {
>>>>>> -error_report("Error get inflight: %d", -ret);
>>>>>> +error_report("Error setting inflight format: %d", -ret);
>>>>>> goto err_guest_notifiers;
>>>>>> }
>>>>>>
>>>>>
Re: [PATCH v2 4/4] vhost-user-scsi: support reconnect to backend
> 2023年7月31日 06:14，Raphael Norwitz 写道：
>
> I don’t think we should be changing any vhost-scsi-common code here. I’d
> rather implement wrappers around vhost_user_scsi_start/stop() around
> vhost_user_scsi_common_start/stop() and check started_vu there.
>
> Otherwise I think this is looking good.
>
> Glad to see you caught the vhost_user_scsi_handle_ouptut and implemented it
> like vhost-user-blk. Can it go in a separate change?
I will fix it in v3.
>
>> On Jul 25, 2023, at 6:42 AM, Li Feng wrote:
>>
>> If the backend crashes and restarts, the device is broken.
>> This patch adds reconnect for vhost-user-scsi.
>>
>> Tested with spdk backend.
>>
>> Signed-off-by: Li Feng
>> ---
>> hw/scsi/vhost-scsi-common.c | 6 +
>> hw/scsi/vhost-user-scsi.c | 220 +++---
>> include/hw/virtio/vhost-scsi-common.h | 3 +
>> include/hw/virtio/vhost-user-scsi.h | 3 +
>> 4 files changed, 211 insertions(+), 21 deletions(-)
>>
>> diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
>> index 664adb15b4..3fde477eee 100644
>> --- a/hw/scsi/vhost-scsi-common.c
>> +++ b/hw/scsi/vhost-scsi-common.c
>> @@ -81,6 +81,7 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>>error_report("Error start vhost dev");
>>goto err_guest_notifiers;
>>}
>> +vsc->started_vu = true;
>>
>>/* guest_notifier_mask/pending not used yet, so just unmask
>> * everything here. virtio-pci will do the right thing by
>> @@ -106,6 +107,11 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
>>VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
>>int ret = 0;
>>
>> +if (!vsc->started_vu) {
>> +return;
>> +}
>> +vsc->started_vu = false;
>> +
>>vhost_dev_stop(>dev, vdev, true);
>>
>>if (k->set_guest_notifiers) {
>> diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
>> index ee99b19e7a..bd32dcf999 100644
>> --- a/hw/scsi/vhost-user-scsi.c
>> +++ b/hw/scsi/vhost-user-scsi.c
>> @@ -46,20 +46,25 @@ enum VhostUserProtocolFeature {
>> static void vhost_user_scsi_set_status(VirtIODevice *vdev, uint8_t status)
>> {
>>VHostUserSCSI *s = (VHostUserSCSI *)vdev;
>> +DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
>>VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
>> -bool start = (status & VIRTIO_CONFIG_S_DRIVER_OK) && vdev->vm_running;
>> +VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
>> +bool should_start = virtio_device_should_start(vdev, status);
>> +int ret;
>>
>> -if (vhost_dev_is_started(>dev) == start) {
>> +if (!s->connected) {
>>return;
>>}
>>
>> -if (start) {
>> -int ret;
>> +if (vhost_dev_is_started(>dev) == should_start) {
>> +return;
>> +}
>>
>> +if (should_start) {
>>ret = vhost_scsi_common_start(vsc);
>>if (ret < 0) {
>>error_report("unable to start vhost-user-scsi: %s",
>> strerror(-ret));
>> -exit(1);
>> +qemu_chr_fe_disconnect(>conf.chardev);
>>}
>>} else {
>>vhost_scsi_common_stop(vsc);
>> @@ -85,8 +90,160 @@ static void vhost_user_scsi_reset(VirtIODevice *vdev)
>>}
>> }
>>
>> -static void vhost_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
>> +static void vhost_user_scsi_handle_output(VirtIODevice *vdev, VirtQueue *vq)
>> {
>> +VHostUserSCSI *s = (VHostUserSCSI *)vdev;
>> +DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
>> +VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
>> +VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
>> +
>> +Error *local_err = NULL;
>> +int i, ret;
>> +
>> +if (!vdev->start_on_kick) {
>> +return;
>> +}
>> +
>> +if (!s->connected) {
>> +return;
>> +}
>> +
>> +if (vhost_dev_is_started(>dev)) {
>> +return;
>> +}
>> +
>> +/*
>> + * Some guests kick before setting VIRTIO_CONFIG_S_DRIVER_OK so start
>> + * vhost here instead of waiting for .set_status().
>> + */
>> +ret = vhost_scsi_common_start(vsc);
>> +if (ret < 0) {
>> +error_reportf_err(local_err, "vhost-user-blk: vhost start failed:
>> ");
&
Re: [PATCH] vhost-user-scsi: support reconnect to backend
Thanks for your reply.
> 2023年7月28日 上午5:21，Raphael Norwitz 写道：
>
>
>
>> On Jul 25, 2023, at 6:19 AM, Li Feng wrote:
>>
>> Thanks for your comments.
>>
>>> 2023年7月25日 上午1:21，Raphael Norwitz 写道：
>>>
>>> Very excited to see this. High level looks good modulo a few small things.
>>>
>>> My major concern is around existing vhost-user-scsi backends which don’t
>>> support VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD. IMO we should hide the
>>> reconnect behavior behind a VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD check. We
>>> may want to do the same for vhost-user-blk.
>>>
>>> The question is then what happens if the check is false. IIUC without an
>>> inflight FD, if a device processes requests out of order, it’s not safe to
>>> continue execution on reconnect, as there’s no way for the backend to know
>>> how to replay IO. Should we permanently wedge the device or have QEMU fail
>>> out? May be nice to have a toggle for this.
>>
>> Based on what MST said, is there anything else I need to do?
>
> I don’t think so.
>
>>>
>>>> On Jul 21, 2023, at 6:51 AM, Li Feng wrote:
>>>>
>>>> If the backend crashes and restarts, the device is broken.
>>>> This patch adds reconnect for vhost-user-scsi.
>>>>
>>>> Tested with spdk backend.
>>>>
>>>> Signed-off-by: Li Feng
>>>> ---
>>>> hw/block/vhost-user-blk.c | 2 -
>>>> hw/scsi/vhost-scsi-common.c | 27 ++---
>>>> hw/scsi/vhost-user-scsi.c | 163 +---
>>>> include/hw/virtio/vhost-user-scsi.h | 3 +
>>>> include/hw/virtio/vhost.h | 2 +
>>>> 5 files changed, 165 insertions(+), 32 deletions(-)
>>>>
>>>> diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
>>>> index eecf3f7a81..f250c740b5 100644
>>>> --- a/hw/block/vhost-user-blk.c
>>>> +++ b/hw/block/vhost-user-blk.c
>>>> @@ -32,8 +32,6 @@
>>>> #include "sysemu/sysemu.h"
>>>> #include "sysemu/runstate.h"
>>>>
>>>> -#define REALIZE_CONNECTION_RETRIES 3
>>>> -
>>>> static const int user_feature_bits[] = {
>>>> VIRTIO_BLK_F_SIZE_MAX,
>>>> VIRTIO_BLK_F_SEG_MAX,
>>>> diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
>>>
>>> Why can’t all the vhost-scsi-common stuff be moved to a separate change?
>>
>> I will move this code to separate patch.
>>>
>>> Especially the stuff introduced for vhost-user-blk in
>>> 1b0063b3048af65dfaae6422a572c87db8575a92 should be moved out.
>> OK.
>>
>>>
>>>> index a06f01af26..08801886b8 100644
>>>> --- a/hw/scsi/vhost-scsi-common.c
>>>> +++ b/hw/scsi/vhost-scsi-common.c
>>>> @@ -52,16 +52,22 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>>>>
>>>> vsc->dev.acked_features = vdev->guest_features;
>>>>
>>>> -assert(vsc->inflight == NULL);
>>>> -vsc->inflight = g_new0(struct vhost_inflight, 1);
>>>> -ret = vhost_dev_get_inflight(>dev,
>>>> - vs->conf.virtqueue_size,
>>>> - vsc->inflight);
>>>> +ret = vhost_dev_prepare_inflight(>dev, vdev);
>>>> if (ret < 0) {
>>>> -error_report("Error get inflight: %d", -ret);
>>>> +error_report("Error setting inflight format: %d", -ret);
>>>> goto err_guest_notifiers;
>>>> }
>>>>
>>>> +if (!vsc->inflight->addr) {
>>>> +ret = vhost_dev_get_inflight(>dev,
>>>> +vs->conf.virtqueue_size,
>>>> +vsc->inflight);
>>>> +if (ret < 0) {
>>>> +error_report("Error get inflight: %d", -ret);
>>>> +goto err_guest_notifiers;
>>>> +}
>>>> +}
>>>> +
>>>> ret = vhost_dev_set_inflight(>dev, vsc->inflight);
>>>> if (ret < 0) {
>>>> error_report("Error set inflight: %d", -ret);
>>>> @@ -85,9 +91,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>>>> return ret;
>>>>
&
Re: [PATCH v2 2/4] vhost-user-common: send get_inflight_fd once
> 2023年7月28日 下午2:04，Michael S. Tsirkin 写道：
>
> On Tue, Jul 25, 2023 at 06:42:45PM +0800, Li Feng wrote:
>> Get_inflight_fd is sent only once. When reconnecting to the backend,
>> qemu sent set_inflight_fd to the backend.
>
> I don't understand what you are trying to say here.
> Should be:
> Currently ABCD. This is wrong/unnecessary because EFG. This patch HIJ.
Thanks, I will reorganize the commit message in v3.
>
>> Signed-off-by: Li Feng
>> ---
>> hw/scsi/vhost-scsi-common.c | 37 ++---
>> 1 file changed, 18 insertions(+), 19 deletions(-)
>>
>> diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
>> index a06f01af26..664adb15b4 100644
>> --- a/hw/scsi/vhost-scsi-common.c
>> +++ b/hw/scsi/vhost-scsi-common.c
>> @@ -52,20 +52,28 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>>
>> vsc->dev.acked_features = vdev->guest_features;
>>
>> -assert(vsc->inflight == NULL);
>> -vsc->inflight = g_new0(struct vhost_inflight, 1);
>> -ret = vhost_dev_get_inflight(>dev,
>> - vs->conf.virtqueue_size,
>> - vsc->inflight);
>> +ret = vhost_dev_prepare_inflight(>dev, vdev);
>> if (ret < 0) {
>> -error_report("Error get inflight: %d", -ret);
>> +error_report("Error setting inflight format: %d", -ret);
>> goto err_guest_notifiers;
>> }
>>
>> -ret = vhost_dev_set_inflight(>dev, vsc->inflight);
>> -if (ret < 0) {
>> -error_report("Error set inflight: %d", -ret);
>> -goto err_guest_notifiers;
>> +if (vsc->inflight) {
>> +if (!vsc->inflight->addr) {
>> +ret = vhost_dev_get_inflight(>dev,
>> +vs->conf.virtqueue_size,
>> +vsc->inflight);
>> +if (ret < 0) {
>> +error_report("Error get inflight: %d", -ret);
>
> As long as you are fixing this - should be "getting inflight”.
I will fix it in v3.
>
>> +goto err_guest_notifiers;
>> +}
>> +}
>> +
>> +ret = vhost_dev_set_inflight(>dev, vsc->inflight);
>> +if (ret < 0) {
>> +error_report("Error set inflight: %d", -ret);
>> +goto err_guest_notifiers;
>> +}
>> }
>>
>> ret = vhost_dev_start(>dev, vdev, true);
>> @@ -85,9 +93,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>> return ret;
>>
>> err_guest_notifiers:
>> -g_free(vsc->inflight);
>> -vsc->inflight = NULL;
>> -
>> k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
>> err_host_notifiers:
>> vhost_dev_disable_notifiers(>dev, vdev);
>> @@ -111,12 +116,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
>> }
>> assert(ret >= 0);
>>
>> -if (vsc->inflight) {
>> -vhost_dev_free_inflight(vsc->inflight);
>> -g_free(vsc->inflight);
>> -vsc->inflight = NULL;
>> -}
>> -
>> vhost_dev_disable_notifiers(>dev, vdev);
>> }
>>
>> --
>> 2.41.0
Re: [PATCH v2 4/4] vhost-user-scsi: support reconnect to backend
> 2023年7月25日 下午6:42，Li Feng 写道：
>
> If the backend crashes and restarts, the device is broken.
> This patch adds reconnect for vhost-user-scsi.
>
> Tested with spdk backend.
>
> Signed-off-by: Li Feng
> ---
> hw/scsi/vhost-scsi-common.c | 6 +
> hw/scsi/vhost-user-scsi.c | 220 +++---
> include/hw/virtio/vhost-scsi-common.h | 3 +
> include/hw/virtio/vhost-user-scsi.h | 3 +
> 4 files changed, 211 insertions(+), 21 deletions(-)
>
> diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
> index 664adb15b4..3fde477eee 100644
> --- a/hw/scsi/vhost-scsi-common.c
> +++ b/hw/scsi/vhost-scsi-common.c
> @@ -81,6 +81,7 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
> error_report("Error start vhost dev");
> goto err_guest_notifiers;
> }
> +vsc->started_vu = true;
>
> /* guest_notifier_mask/pending not used yet, so just unmask
> * everything here. virtio-pci will do the right thing by
> @@ -106,6 +107,11 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
> VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
> int ret = 0;
>
> +if (!vsc->started_vu) {
> +return;
> +}
> +vsc->started_vu = false;
> +
> vhost_dev_stop(>dev, vdev, true);
>
> if (k->set_guest_notifiers) {
> diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
> index ee99b19e7a..bd32dcf999 100644
> --- a/hw/scsi/vhost-user-scsi.c
> +++ b/hw/scsi/vhost-user-scsi.c
> @@ -46,20 +46,25 @@ enum VhostUserProtocolFeature {
> static void vhost_user_scsi_set_status(VirtIODevice *vdev, uint8_t status)
> {
> VHostUserSCSI *s = (VHostUserSCSI *)vdev;
> +DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
> VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
> -bool start = (status & VIRTIO_CONFIG_S_DRIVER_OK) && vdev->vm_running;
> +VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
> +bool should_start = virtio_device_should_start(vdev, status);
> +int ret;
>
> -if (vhost_dev_is_started(>dev) == start) {
> +if (!s->connected) {
> return;
> }
>
> -if (start) {
> -int ret;
> +if (vhost_dev_is_started(>dev) == should_start) {
> +return;
> +}
>
> +if (should_start) {
> ret = vhost_scsi_common_start(vsc);
> if (ret < 0) {
> error_report("unable to start vhost-user-scsi: %s",
> strerror(-ret));
> -exit(1);
> +qemu_chr_fe_disconnect(>conf.chardev);
> }
> } else {
> vhost_scsi_common_stop(vsc);
> @@ -85,8 +90,160 @@ static void vhost_user_scsi_reset(VirtIODevice *vdev)
> }
> }
>
> -static void vhost_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
> +static void vhost_user_scsi_handle_output(VirtIODevice *vdev, VirtQueue *vq)
> {
> +VHostUserSCSI *s = (VHostUserSCSI *)vdev;
> +DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
> +VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
> +VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
> +
> +Error *local_err = NULL;
> +int i, ret;
> +
> +if (!vdev->start_on_kick) {
> +return;
> +}
> +
> +if (!s->connected) {
> +return;
> +}
> +
> +if (vhost_dev_is_started(>dev)) {
> +return;
> +}
> +
> +/*
> + * Some guests kick before setting VIRTIO_CONFIG_S_DRIVER_OK so start
> + * vhost here instead of waiting for .set_status().
> + */
> +ret = vhost_scsi_common_start(vsc);
> +if (ret < 0) {
> +error_reportf_err(local_err, "vhost-user-blk: vhost start failed: “);
Need fix typo in v3. s/vhost-user-blk/vhost-user-scsi/g
> +qemu_chr_fe_disconnect(>conf.chardev);
> +return;
> +}
> +
> +/* Kick right away to begin processing requests already in vring */
> +for (i = 0; i < vsc->dev.nvqs; i++) {
> +VirtQueue *kick_vq = virtio_get_queue(vdev, i);
> +
> +if (!virtio_queue_get_desc_addr(vdev, i)) {
> +continue;
> +}
> +event_notifier_set(virtio_queue_get_host_notifier(kick_vq));
> +}
> +}
> +
> +static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
> +{
> +VirtIODevice *vdev = VIRTIO_DEVICE(dev);
> +VHostUserSCSI *s = VHOST_USER_SCSI(vdev);
> +VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
> +VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
> +int ret = 0;
> +
> +if (s->connected) {
> +return 0;
> +
[PATCH v2 3/4] vhost: move and rename the conn retry times
Multile devices need this macro, move it to a common header.
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 4 +---
hw/virtio/vhost-user-gpio.c | 3 +--
include/hw/virtio/vhost.h | 2 ++
3 files changed, 4 insertions(+), 5 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index eecf3f7a81..3c69fa47d5 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -32,8 +32,6 @@
#include "sysemu/sysemu.h"
#include "sysemu/runstate.h"
-#define REALIZE_CONNECTION_RETRIES 3
-
static const int user_feature_bits[] = {
VIRTIO_BLK_F_SIZE_MAX,
VIRTIO_BLK_F_SEG_MAX,
@@ -482,7 +480,7 @@ static void vhost_user_blk_device_realize(DeviceState *dev,
Error **errp)
s->inflight = g_new0(struct vhost_inflight, 1);
s->vhost_vqs = g_new0(struct vhost_virtqueue, s->num_queues);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
assert(!*errp);
do {
if (*errp) {
diff --git a/hw/virtio/vhost-user-gpio.c b/hw/virtio/vhost-user-gpio.c
index 3b013f2d0f..d9979aa5db 100644
--- a/hw/virtio/vhost-user-gpio.c
+++ b/hw/virtio/vhost-user-gpio.c
@@ -15,7 +15,6 @@
#include "standard-headers/linux/virtio_ids.h"
#include "trace.h"
-#define REALIZE_CONNECTION_RETRIES 3
#define VHOST_NVQS 2
/* Features required from VirtIO */
@@ -359,7 +358,7 @@ static void vu_gpio_device_realize(DeviceState *dev, Error
**errp)
qemu_chr_fe_set_handlers(>chardev, NULL, NULL, vu_gpio_event, NULL,
dev, NULL, true);
-retries = REALIZE_CONNECTION_RETRIES;
+retries = VU_REALIZE_CONN_RETRIES;
g_assert(!*errp);
do {
if (*errp) {
diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h
index 6a173cb9fa..ca3131b1af 100644
--- a/include/hw/virtio/vhost.h
+++ b/include/hw/virtio/vhost.h
@@ -8,6 +8,8 @@
#define VHOST_F_DEVICE_IOTLB 63
#define VHOST_USER_F_PROTOCOL_FEATURES 30
+#define VU_REALIZE_CONN_RETRIES 3
+
/* Generic structures common for any vhost based device. */
struct vhost_inflight {
--
2.41.0
[PATCH v2 1/4] vhost: fix the fd leak
When the vhost-user reconnect to the backend, the notifer should be
cleanup. Otherwise, the fd resource will be exhausted.
Fixes: f9a09ca3ea ("vhost: add support for configure interrupt")
Signed-off-by: Li Feng
---
hw/virtio/vhost.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c
index abf0d03c8d..e2f6ffb446 100644
--- a/hw/virtio/vhost.c
+++ b/hw/virtio/vhost.c
@@ -2044,6 +2044,8 @@ void vhost_dev_stop(struct vhost_dev *hdev, VirtIODevice
*vdev, bool vrings)
event_notifier_test_and_clear(
>vqs[VHOST_QUEUE_NUM_CONFIG_INR].masked_config_notifier);
event_notifier_test_and_clear(>config_notifier);
+event_notifier_cleanup(
+>vqs[VHOST_QUEUE_NUM_CONFIG_INR].masked_config_notifier);
trace_vhost_dev_stop(hdev, vdev->name, vrings);
--
2.41.0
[PATCH v2 0/4] Implement reconnect for vhost-user-scsi
Hi, This patchset adds reconnect support for vhost-user-scsi. At the same times, fix vhost fd leak and refactor some code. Changes for v2: - Split the v1 patch to small separate patchset; - New patch for fixing fd leak, which has sent to reviewers in another mail; - Implement the `vhost_user_scsi_handle_output`; - Add the started_vu safe check; - Fix error handler; - Check the inflight before set/get inflight fd. Li Feng (4): vhost: fix the fd leak vhost-user-common: send get_inflight_fd once vhost: move and rename the conn retry times vhost-user-scsi: support reconnect to backend hw/block/vhost-user-blk.c | 4 +- hw/scsi/vhost-scsi-common.c | 43 ++--- hw/scsi/vhost-user-scsi.c | 220 +++--- hw/virtio/vhost-user-gpio.c | 3 +- hw/virtio/vhost.c | 2 + include/hw/virtio/vhost-scsi-common.h | 3 + include/hw/virtio/vhost-user-scsi.h | 3 + include/hw/virtio/vhost.h | 2 + 8 files changed, 235 insertions(+), 45 deletions(-) -- 2.41.0
[PATCH v2 4/4] vhost-user-scsi: support reconnect to backend
If the backend crashes and restarts, the device is broken.
This patch adds reconnect for vhost-user-scsi.
Tested with spdk backend.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 6 +
hw/scsi/vhost-user-scsi.c | 220 +++---
include/hw/virtio/vhost-scsi-common.h | 3 +
include/hw/virtio/vhost-user-scsi.h | 3 +
4 files changed, 211 insertions(+), 21 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index 664adb15b4..3fde477eee 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -81,6 +81,7 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
error_report("Error start vhost dev");
goto err_guest_notifiers;
}
+vsc->started_vu = true;
/* guest_notifier_mask/pending not used yet, so just unmask
* everything here. virtio-pci will do the right thing by
@@ -106,6 +107,11 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
int ret = 0;
+if (!vsc->started_vu) {
+return;
+}
+vsc->started_vu = false;
+
vhost_dev_stop(>dev, vdev, true);
if (k->set_guest_notifiers) {
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index ee99b19e7a..bd32dcf999 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -46,20 +46,25 @@ enum VhostUserProtocolFeature {
static void vhost_user_scsi_set_status(VirtIODevice *vdev, uint8_t status)
{
VHostUserSCSI *s = (VHostUserSCSI *)vdev;
+DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
-bool start = (status & VIRTIO_CONFIG_S_DRIVER_OK) && vdev->vm_running;
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+bool should_start = virtio_device_should_start(vdev, status);
+int ret;
-if (vhost_dev_is_started(>dev) == start) {
+if (!s->connected) {
return;
}
-if (start) {
-int ret;
+if (vhost_dev_is_started(>dev) == should_start) {
+return;
+}
+if (should_start) {
ret = vhost_scsi_common_start(vsc);
if (ret < 0) {
error_report("unable to start vhost-user-scsi: %s",
strerror(-ret));
-exit(1);
+qemu_chr_fe_disconnect(>conf.chardev);
}
} else {
vhost_scsi_common_stop(vsc);
@@ -85,8 +90,160 @@ static void vhost_user_scsi_reset(VirtIODevice *vdev)
}
}
-static void vhost_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
+static void vhost_user_scsi_handle_output(VirtIODevice *vdev, VirtQueue *vq)
{
+VHostUserSCSI *s = (VHostUserSCSI *)vdev;
+DeviceState *dev = >parent_obj.parent_obj.parent_obj.parent_obj;
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+
+Error *local_err = NULL;
+int i, ret;
+
+if (!vdev->start_on_kick) {
+return;
+}
+
+if (!s->connected) {
+return;
+}
+
+if (vhost_dev_is_started(>dev)) {
+return;
+}
+
+/*
+ * Some guests kick before setting VIRTIO_CONFIG_S_DRIVER_OK so start
+ * vhost here instead of waiting for .set_status().
+ */
+ret = vhost_scsi_common_start(vsc);
+if (ret < 0) {
+error_reportf_err(local_err, "vhost-user-blk: vhost start failed: ");
+qemu_chr_fe_disconnect(>conf.chardev);
+return;
+}
+
+/* Kick right away to begin processing requests already in vring */
+for (i = 0; i < vsc->dev.nvqs; i++) {
+VirtQueue *kick_vq = virtio_get_queue(vdev, i);
+
+if (!virtio_queue_get_desc_addr(vdev, i)) {
+continue;
+}
+event_notifier_set(virtio_queue_get_host_notifier(kick_vq));
+}
+}
+
+static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
+{
+VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+VHostUserSCSI *s = VHOST_USER_SCSI(vdev);
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+int ret = 0;
+
+if (s->connected) {
+return 0;
+}
+s->connected = true;
+
+vsc->dev.num_queues = vs->conf.num_queues;
+vsc->dev.nvqs = VIRTIO_SCSI_VQ_NUM_FIXED + vs->conf.num_queues;
+vsc->dev.vqs = s->vhost_vqs;
+vsc->dev.vq_index = 0;
+vsc->dev.backend_features = 0;
+
+ret = vhost_dev_init(>dev, >vhost_user, VHOST_BACKEND_TYPE_USER, 0,
+ errp);
+if (ret < 0) {
+return ret;
+}
+
+/* restore vhost state */
+if (virtio_device_started(vdev, vdev->status)) {
+ret = vhost_scsi_common_start(vsc);
+if (ret < 0) {
+return ret;
+}
+}
+
+return 0;
+}
+
+static void vhost_user_scsi_event(void *opaque, QEMU
[PATCH v2 2/4] vhost-user-common: send get_inflight_fd once
Get_inflight_fd is sent only once. When reconnecting to the backend,
qemu sent set_inflight_fd to the backend.
Signed-off-by: Li Feng
---
hw/scsi/vhost-scsi-common.c | 37 ++---
1 file changed, 18 insertions(+), 19 deletions(-)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a06f01af26..664adb15b4 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -52,20 +52,28 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vsc->dev.acked_features = vdev->guest_features;
-assert(vsc->inflight == NULL);
-vsc->inflight = g_new0(struct vhost_inflight, 1);
-ret = vhost_dev_get_inflight(>dev,
- vs->conf.virtqueue_size,
- vsc->inflight);
+ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error get inflight: %d", -ret);
+error_report("Error setting inflight format: %d", -ret);
goto err_guest_notifiers;
}
-ret = vhost_dev_set_inflight(>dev, vsc->inflight);
-if (ret < 0) {
-error_report("Error set inflight: %d", -ret);
-goto err_guest_notifiers;
+if (vsc->inflight) {
+if (!vsc->inflight->addr) {
+ret = vhost_dev_get_inflight(>dev,
+vs->conf.virtqueue_size,
+vsc->inflight);
+if (ret < 0) {
+error_report("Error get inflight: %d", -ret);
+goto err_guest_notifiers;
+}
+}
+
+ret = vhost_dev_set_inflight(>dev, vsc->inflight);
+if (ret < 0) {
+error_report("Error set inflight: %d", -ret);
+goto err_guest_notifiers;
+}
}
ret = vhost_dev_start(>dev, vdev, true);
@@ -85,9 +93,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
return ret;
err_guest_notifiers:
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-
k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
err_host_notifiers:
vhost_dev_disable_notifiers(>dev, vdev);
@@ -111,12 +116,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
}
assert(ret >= 0);
-if (vsc->inflight) {
-vhost_dev_free_inflight(vsc->inflight);
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-}
-
vhost_dev_disable_notifiers(>dev, vdev);
}
--
2.41.0
Re: [PATCH] vhost-user-scsi: support reconnect to backend
Thanks for your comments.
> 2023年7月25日 上午1:21，Raphael Norwitz 写道：
>
> Very excited to see this. High level looks good modulo a few small things.
>
> My major concern is around existing vhost-user-scsi backends which don’t
> support VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD. IMO we should hide the
> reconnect behavior behind a VHOST_USER_PROTOCOL_F_INFLIGHT_SHMFD check. We
> may want to do the same for vhost-user-blk.
>
> The question is then what happens if the check is false. IIUC without an
> inflight FD, if a device processes requests out of order, it’s not safe to
> continue execution on reconnect, as there’s no way for the backend to know
> how to replay IO. Should we permanently wedge the device or have QEMU fail
> out? May be nice to have a toggle for this.
Based on what MST said, is there anything else I need to do?
>
>> On Jul 21, 2023, at 6:51 AM, Li Feng wrote:
>>
>> If the backend crashes and restarts, the device is broken.
>> This patch adds reconnect for vhost-user-scsi.
>>
>> Tested with spdk backend.
>>
>> Signed-off-by: Li Feng
>> ---
>> hw/block/vhost-user-blk.c | 2 -
>> hw/scsi/vhost-scsi-common.c | 27 ++---
>> hw/scsi/vhost-user-scsi.c | 163 +---
>> include/hw/virtio/vhost-user-scsi.h | 3 +
>> include/hw/virtio/vhost.h | 2 +
>> 5 files changed, 165 insertions(+), 32 deletions(-)
>>
>> diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
>> index eecf3f7a81..f250c740b5 100644
>> --- a/hw/block/vhost-user-blk.c
>> +++ b/hw/block/vhost-user-blk.c
>> @@ -32,8 +32,6 @@
>> #include "sysemu/sysemu.h"
>> #include "sysemu/runstate.h"
>>
>> -#define REALIZE_CONNECTION_RETRIES 3
>> -
>> static const int user_feature_bits[] = {
>>VIRTIO_BLK_F_SIZE_MAX,
>>VIRTIO_BLK_F_SEG_MAX,
>> diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
>
> Why can’t all the vhost-scsi-common stuff be moved to a separate change?
I will move this code to separate patch.
>
> Especially the stuff introduced for vhost-user-blk in
> 1b0063b3048af65dfaae6422a572c87db8575a92 should be moved out.
OK.
>
>> index a06f01af26..08801886b8 100644
>> --- a/hw/scsi/vhost-scsi-common.c
>> +++ b/hw/scsi/vhost-scsi-common.c
>> @@ -52,16 +52,22 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>>
>>vsc->dev.acked_features = vdev->guest_features;
>>
>> -assert(vsc->inflight == NULL);
>> -vsc->inflight = g_new0(struct vhost_inflight, 1);
>> -ret = vhost_dev_get_inflight(>dev,
>> - vs->conf.virtqueue_size,
>> - vsc->inflight);
>> +ret = vhost_dev_prepare_inflight(>dev, vdev);
>>if (ret < 0) {
>> -error_report("Error get inflight: %d", -ret);
>> +error_report("Error setting inflight format: %d", -ret);
>>goto err_guest_notifiers;
>>}
>>
>> +if (!vsc->inflight->addr) {
>> +ret = vhost_dev_get_inflight(>dev,
>> +vs->conf.virtqueue_size,
>> +vsc->inflight);
>> +if (ret < 0) {
>> +error_report("Error get inflight: %d", -ret);
>> +goto err_guest_notifiers;
>> +}
>> +}
>> +
>>ret = vhost_dev_set_inflight(>dev, vsc->inflight);
>>if (ret < 0) {
>>error_report("Error set inflight: %d", -ret);
>> @@ -85,9 +91,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
>>return ret;
>>
>> err_guest_notifiers:
>> -g_free(vsc->inflight);
>> -vsc->inflight = NULL;
>> -
>>k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
>> err_host_notifiers:
>>vhost_dev_disable_notifiers(>dev, vdev);
>> @@ -111,12 +114,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
>>}
>>assert(ret >= 0);
>>
>
> In the vhost-scsi (kernel backend) path, what will cleanup vsc->inflight now?
OK, we should check the vsc->inflight if it is null, the vhost-scsi doesn’t
allocate the
inflight object memory.
>
>> -if (vsc->inflight) {
>> -vhost_dev_free_inflight(vsc->inflight);
>> -g_free(vsc->inflight);
>> -vsc->inflight = NULL;
>> -}
>> -
>>vhost_dev_disable_notifiers(>dev, vdev);
>> }
>>
>> d
[PATCH] vhost-user-scsi: support reconnect to backend
If the backend crashes and restarts, the device is broken.
This patch adds reconnect for vhost-user-scsi.
Tested with spdk backend.
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 2 -
hw/scsi/vhost-scsi-common.c | 27 ++---
hw/scsi/vhost-user-scsi.c | 163 +---
include/hw/virtio/vhost-user-scsi.h | 3 +
include/hw/virtio/vhost.h | 2 +
5 files changed, 165 insertions(+), 32 deletions(-)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index eecf3f7a81..f250c740b5 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -32,8 +32,6 @@
#include "sysemu/sysemu.h"
#include "sysemu/runstate.h"
-#define REALIZE_CONNECTION_RETRIES 3
-
static const int user_feature_bits[] = {
VIRTIO_BLK_F_SIZE_MAX,
VIRTIO_BLK_F_SEG_MAX,
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index a06f01af26..08801886b8 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -52,16 +52,22 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
vsc->dev.acked_features = vdev->guest_features;
-assert(vsc->inflight == NULL);
-vsc->inflight = g_new0(struct vhost_inflight, 1);
-ret = vhost_dev_get_inflight(>dev,
- vs->conf.virtqueue_size,
- vsc->inflight);
+ret = vhost_dev_prepare_inflight(>dev, vdev);
if (ret < 0) {
-error_report("Error get inflight: %d", -ret);
+error_report("Error setting inflight format: %d", -ret);
goto err_guest_notifiers;
}
+if (!vsc->inflight->addr) {
+ret = vhost_dev_get_inflight(>dev,
+vs->conf.virtqueue_size,
+vsc->inflight);
+if (ret < 0) {
+error_report("Error get inflight: %d", -ret);
+goto err_guest_notifiers;
+}
+}
+
ret = vhost_dev_set_inflight(>dev, vsc->inflight);
if (ret < 0) {
error_report("Error set inflight: %d", -ret);
@@ -85,9 +91,6 @@ int vhost_scsi_common_start(VHostSCSICommon *vsc)
return ret;
err_guest_notifiers:
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-
k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
err_host_notifiers:
vhost_dev_disable_notifiers(>dev, vdev);
@@ -111,12 +114,6 @@ void vhost_scsi_common_stop(VHostSCSICommon *vsc)
}
assert(ret >= 0);
-if (vsc->inflight) {
-vhost_dev_free_inflight(vsc->inflight);
-g_free(vsc->inflight);
-vsc->inflight = NULL;
-}
-
vhost_dev_disable_notifiers(>dev, vdev);
}
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index ee99b19e7a..e0e88b0c42 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -89,14 +89,126 @@ static void vhost_dummy_handle_output(VirtIODevice *vdev,
VirtQueue *vq)
{
}
+static int vhost_user_scsi_connect(DeviceState *dev, Error **errp)
+{
+VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+VHostUserSCSI *s = VHOST_USER_SCSI(vdev);
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+int ret = 0;
+
+if (s->connected) {
+return 0;
+}
+s->connected = true;
+
+vsc->dev.num_queues = vs->conf.num_queues;
+vsc->dev.nvqs = VIRTIO_SCSI_VQ_NUM_FIXED + vs->conf.num_queues;
+vsc->dev.vqs = s->vhost_vqs;
+vsc->dev.vq_index = 0;
+vsc->dev.backend_features = 0;
+
+ret = vhost_dev_init(>dev, >vhost_user, VHOST_BACKEND_TYPE_USER, 0,
+ errp);
+if (ret < 0) {
+return ret;
+}
+
+/* restore vhost state */
+if (virtio_device_started(vdev, vdev->status)) {
+ret = vhost_scsi_common_start(vsc);
+if (ret < 0) {
+return ret;
+}
+}
+
+return 0;
+}
+
+static void vhost_user_scsi_event(void *opaque, QEMUChrEvent event);
+
+static void vhost_user_scsi_disconnect(DeviceState *dev)
+{
+VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+VHostUserSCSI *s = VHOST_USER_SCSI(vdev);
+VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
+VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(dev);
+
+if (!s->connected) {
+return;
+}
+s->connected = false;
+
+vhost_scsi_common_stop(vsc);
+
+vhost_dev_cleanup(>dev);
+
+/* Re-instate the event handler for new connections */
+qemu_chr_fe_set_handlers(>conf.chardev, NULL, NULL,
+ vhost_user_scsi_event, NULL, dev, NULL, true);
+}
+
+static void vhost_user_scsi_event(void *opaque, QEMUChrEvent event)
+{
+DeviceState *dev = opaque;
+VirtIODevice *vdev = VIRTIO_DEVICE(dev);
+VHostUserSCSI *s =
[PATCH] vhost-user-blk: fix the resize crash
If the os is not installed and doesn't have the virtio guest driver,
the vhost dev isn't started, so the dev->vdev is NULL.
Reproduce: mount a Win 2019 iso, go into the install ui, then resize
the virtio-blk device, qemu crash.
Signed-off-by: Li Feng
---
hw/block/vhost-user-blk.c | 4
1 file changed, 4 insertions(+)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 9117222456..db30bb754f 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -95,6 +95,10 @@ static int vhost_user_blk_handle_config_change(struct
vhost_dev *dev)
VHostUserBlk *s = VHOST_USER_BLK(dev->vdev);
Error *local_err = NULL;
+if (!dev->started) {
+return 0;
+}
+
ret = vhost_dev_get_config(dev, (uint8_t *),
sizeof(struct virtio_blk_config),
_err);
--
2.37.3
[PATCH v5] file-posix: detect the lock using the real file
This patch addresses this issue:
When accessing a volume on an NFS filesystem without supporting the file lock,
tools, like qemu-img, will complain "Failed to lock byte 100".
In the original code, the qemu_has_ofd_lock will test the lock on the
"/dev/null" pseudo-file. Actually, the file.locking is per-drive property,
which depends on the underlay filesystem.
In this patch, add a new 'qemu_has_file_lock' to detect whether the
file supports the file lock. And disable the lock when the underlay file
system doesn't support locks.
Signed-off-by: Li Feng
---
v5: simplify the code.
v4: use the fd as the qemu_has_file_lock argument.
v3: don't call the qemu_has_ofd_lock, use a new function instead.
v2: remove the refactoring.
---
block/file-posix.c | 61 +++-
include/qemu/osdep.h | 1 +
util/osdep.c | 14 ++
3 files changed, 47 insertions(+), 29 deletions(-)
diff --git a/block/file-posix.c b/block/file-posix.c
index 806764f7e3..4e00111031 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -584,6 +584,21 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
s->use_linux_io_uring = (aio == BLOCKDEV_AIO_OPTIONS_IO_URING);
#endif
+s->open_flags = open_flags;
+raw_parse_flags(bdrv_flags, >open_flags, false);
+
+s->fd = -1;
+fd = qemu_open(filename, s->open_flags, errp);
+ret = fd < 0 ? -errno : 0;
+
+if (ret < 0) {
+if (ret == -EROFS) {
+ret = -EACCES;
+}
+goto fail;
+}
+s->fd = fd;
+
locking = qapi_enum_parse(_lookup,
qemu_opt_get(opts, "locking"),
ON_OFF_AUTO_AUTO, _err);
@@ -606,7 +621,7 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
s->use_lock = false;
break;
case ON_OFF_AUTO_AUTO:
-s->use_lock = qemu_has_ofd_lock();
+s->use_lock = qemu_has_file_lock(s->fd) && qemu_has_ofd_lock();
break;
default:
abort();
@@ -625,22 +640,6 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
s->drop_cache = qemu_opt_get_bool(opts, "drop-cache", true);
s->check_cache_dropped = qemu_opt_get_bool(opts, "x-check-cache-dropped",
false);
-
-s->open_flags = open_flags;
-raw_parse_flags(bdrv_flags, >open_flags, false);
-
-s->fd = -1;
-fd = qemu_open(filename, s->open_flags, errp);
-ret = fd < 0 ? -errno : 0;
-
-if (ret < 0) {
-if (ret == -EROFS) {
-ret = -EACCES;
-}
-goto fail;
-}
-s->fd = fd;
-
/* Check s->open_flags rather than bdrv_flags due to auto-read-only */
if (s->open_flags & O_RDWR) {
ret = check_hdev_writable(s->fd);
@@ -2388,6 +2387,7 @@ raw_co_create(BlockdevCreateOptions *options, Error
**errp)
int fd;
uint64_t perm, shared;
int result = 0;
+bool use_lock;
/* Validate options and set default values */
assert(options->driver == BLOCKDEV_DRIVER_FILE);
@@ -2428,19 +2428,22 @@ raw_co_create(BlockdevCreateOptions *options, Error
**errp)
perm = BLK_PERM_WRITE | BLK_PERM_RESIZE;
shared = BLK_PERM_ALL & ~BLK_PERM_RESIZE;
-/* Step one: Take locks */
-result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
-if (result < 0) {
-goto out_close;
-}
+use_lock = qemu_has_file_lock(fd);
+if (use_lock) {
+/* Step one: Take locks */
+result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
+if (result < 0) {
+goto out_close;
+}
-/* Step two: Check that nobody else has taken conflicting locks */
-result = raw_check_lock_bytes(fd, perm, shared, errp);
-if (result < 0) {
-error_append_hint(errp,
- "Is another process using the image [%s]?\n",
- file_opts->filename);
-goto out_unlock;
+/* Step two: Check that nobody else has taken conflicting locks */
+result = raw_check_lock_bytes(fd, perm, shared, errp);
+if (result < 0) {
+error_append_hint(errp,
+ "Is another process using the image [%s]?\n",
+ file_opts->filename);
+goto out_unlock;
+}
}
/* Clear the file by truncating it to 0 */
diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h
index f9ec8c84e9..c7587be99d 100644
--- a/include/qemu/osdep.h
+++ b/include/qemu/osdep.h
@@ -513,6 +513,7 @@ int qemu_lock_fd(int fd, int64_t start, int64_t len, bool
exclusive);
int qemu_unlock_fd(int fd, int64_t start, int64_t len);
int qemu_lock_fd_test(int fd, int64_t start, int64_t len, bool exclusive)
Re: [PATCH v4] file-posix: detect the lock using the real file
Daniel P. Berrangé 于2020年12月15日周二 下午6:08写道：
>
> On Tue, Dec 15, 2020 at 03:09:28PM +0800, Li Feng wrote:
> > This patch addresses this issue:
> > When accessing a volume on an NFS filesystem without supporting the file
> > lock,
> > tools, like qemu-img, will complain "Failed to lock byte 100".
> >
> > In the original code, the qemu_has_ofd_lock will test the lock on the
> > "/dev/null" pseudo-file. Actually, the file.locking is per-drive property,
> > which depends on the underlay filesystem.
> >
> > In this patch, add a new 'qemu_has_file_lock' to detect whether the
> > file supports the file lock. And disable the lock when the underlay file
> > system doesn't support locks.
> >
> > Signed-off-by: Li Feng
> > ---
> > v4: use the fd as the qemu_has_file_lock argument.
> > v3: don't call the qemu_has_ofd_lock, use a new function instead.
> > v2: remove the refactoring.
> > ---
> > block/file-posix.c | 66 +---
> > include/qemu/osdep.h | 1 +
> > util/osdep.c | 19 +
> > 3 files changed, 58 insertions(+), 28 deletions(-)
> >
> > diff --git a/block/file-posix.c b/block/file-posix.c
> > index 806764f7e3..9708212f01 100644
> > --- a/block/file-posix.c
> > +++ b/block/file-posix.c
> > @@ -584,6 +584,21 @@ static int raw_open_common(BlockDriverState *bs, QDict
> > *options,
> > s->use_linux_io_uring = (aio == BLOCKDEV_AIO_OPTIONS_IO_URING);
> > #endif
> >
> > +s->open_flags = open_flags;
> > +raw_parse_flags(bdrv_flags, >open_flags, false);
> > +
> > +s->fd = -1;
> > +fd = qemu_open(filename, s->open_flags, errp);
> > +ret = fd < 0 ? -errno : 0;
> > +
> > +if (ret < 0) {
> > +if (ret == -EROFS) {
> > +ret = -EACCES;
> > +}
> > +goto fail;
> > +}
> > +s->fd = fd;
> > +
> > locking = qapi_enum_parse(_lookup,
> >qemu_opt_get(opts, "locking"),
> >ON_OFF_AUTO_AUTO, _err);
> > @@ -607,6 +622,13 @@ static int raw_open_common(BlockDriverState *bs, QDict
> > *options,
> > break;
> > case ON_OFF_AUTO_AUTO:
> > s->use_lock = qemu_has_ofd_lock();
> > +if (s->use_lock && !qemu_has_file_lock(s->fd)) {
> > +/*
> > + * When the os supports the OFD lock, but the filesystem
> > doesn't
> > + * support, just disable the file lock.
> > + */
> > +s->use_lock = false;
> > +}
>
> This should just be
>
> s->use_lock = qemu_has_file_lock(s->fd) && qemu_has_ofd_lock();
>
Acked.
>
>
> > diff --git a/util/osdep.c b/util/osdep.c
> > index 66d01b9160..07de97e2c5 100644
> > --- a/util/osdep.c
> > +++ b/util/osdep.c
> > @@ -225,6 +225,25 @@ static void qemu_probe_lock_ops(void)
> > }
> > }
> >
> > +bool qemu_has_file_lock(int fd)
> > +{
> > +#ifdef F_OFD_SETLK
> > +int cmd = F_OFD_GETLK;
> > +#else
> > +int cmd = F_GETLK;
> > +#endif
>
> This should unconditionally use F_GETLK.
Acked.
>
> > +int ret;
> > +struct flock fl = {
> > +.l_whence = SEEK_SET,
> > +.l_start = 0,
> > +.l_len= 0,
> > +.l_type = F_WRLCK,
> > +};
> > +
> > +ret = fcntl(fd, cmd, );
> > +return ret == 0;
> > +}
>
> Regards,
> Daniel
> --
> |: https://berrange.com -o-https://www.flickr.com/photos/dberrange :|
> |: https://libvirt.org -o-https://fstop138.berrange.com :|
> |: https://entangle-photo.org-o-https://www.instagram.com/dberrange :|
>
[PATCH v4] file-posix: detect the lock using the real file
This patch addresses this issue:
When accessing a volume on an NFS filesystem without supporting the file lock,
tools, like qemu-img, will complain "Failed to lock byte 100".
In the original code, the qemu_has_ofd_lock will test the lock on the
"/dev/null" pseudo-file. Actually, the file.locking is per-drive property,
which depends on the underlay filesystem.
In this patch, add a new 'qemu_has_file_lock' to detect whether the
file supports the file lock. And disable the lock when the underlay file
system doesn't support locks.
Signed-off-by: Li Feng
---
v4: use the fd as the qemu_has_file_lock argument.
v3: don't call the qemu_has_ofd_lock, use a new function instead.
v2: remove the refactoring.
---
block/file-posix.c | 66 +---
include/qemu/osdep.h | 1 +
util/osdep.c | 19 +
3 files changed, 58 insertions(+), 28 deletions(-)
diff --git a/block/file-posix.c b/block/file-posix.c
index 806764f7e3..9708212f01 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -584,6 +584,21 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
s->use_linux_io_uring = (aio == BLOCKDEV_AIO_OPTIONS_IO_URING);
#endif
+s->open_flags = open_flags;
+raw_parse_flags(bdrv_flags, >open_flags, false);
+
+s->fd = -1;
+fd = qemu_open(filename, s->open_flags, errp);
+ret = fd < 0 ? -errno : 0;
+
+if (ret < 0) {
+if (ret == -EROFS) {
+ret = -EACCES;
+}
+goto fail;
+}
+s->fd = fd;
+
locking = qapi_enum_parse(_lookup,
qemu_opt_get(opts, "locking"),
ON_OFF_AUTO_AUTO, _err);
@@ -607,6 +622,13 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
break;
case ON_OFF_AUTO_AUTO:
s->use_lock = qemu_has_ofd_lock();
+if (s->use_lock && !qemu_has_file_lock(s->fd)) {
+/*
+ * When the os supports the OFD lock, but the filesystem doesn't
+ * support, just disable the file lock.
+ */
+s->use_lock = false;
+}
break;
default:
abort();
@@ -625,22 +647,6 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
s->drop_cache = qemu_opt_get_bool(opts, "drop-cache", true);
s->check_cache_dropped = qemu_opt_get_bool(opts, "x-check-cache-dropped",
false);
-
-s->open_flags = open_flags;
-raw_parse_flags(bdrv_flags, >open_flags, false);
-
-s->fd = -1;
-fd = qemu_open(filename, s->open_flags, errp);
-ret = fd < 0 ? -errno : 0;
-
-if (ret < 0) {
-if (ret == -EROFS) {
-ret = -EACCES;
-}
-goto fail;
-}
-s->fd = fd;
-
/* Check s->open_flags rather than bdrv_flags due to auto-read-only */
if (s->open_flags & O_RDWR) {
ret = check_hdev_writable(s->fd);
@@ -2388,6 +2394,7 @@ raw_co_create(BlockdevCreateOptions *options, Error
**errp)
int fd;
uint64_t perm, shared;
int result = 0;
+bool use_lock;
/* Validate options and set default values */
assert(options->driver == BLOCKDEV_DRIVER_FILE);
@@ -2428,19 +2435,22 @@ raw_co_create(BlockdevCreateOptions *options, Error
**errp)
perm = BLK_PERM_WRITE | BLK_PERM_RESIZE;
shared = BLK_PERM_ALL & ~BLK_PERM_RESIZE;
-/* Step one: Take locks */
-result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
-if (result < 0) {
-goto out_close;
-}
+use_lock = qemu_has_file_lock(fd);
+if (use_lock) {
+/* Step one: Take locks */
+result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
+if (result < 0) {
+goto out_close;
+}
-/* Step two: Check that nobody else has taken conflicting locks */
-result = raw_check_lock_bytes(fd, perm, shared, errp);
-if (result < 0) {
-error_append_hint(errp,
- "Is another process using the image [%s]?\n",
- file_opts->filename);
-goto out_unlock;
+/* Step two: Check that nobody else has taken conflicting locks */
+result = raw_check_lock_bytes(fd, perm, shared, errp);
+if (result < 0) {
+error_append_hint(errp,
+ "Is another process using the image [%s]?\n",
+ file_opts->filename);
+goto out_unlock;
+}
}
/* Clear the file by truncating it to 0 */
diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h
index f9ec8c84e9..c7587be99d 100644
--- a/include/qemu/osdep.h
+++ b/include/qemu/osdep.h
@@ -513,6 +513,7 @@ int qemu_lock_fd(int fd, int64_t start, int64_t len, bool
ex
[PATCH v3] file-posix: detect the lock using the real file
This patch addresses this issue:
When accessing a volume on an NFS filesystem without supporting the file lock,
tools, like qemu-img, will complain "Failed to lock byte 100".
Add a new function 'qemu_has_file_lock' to detect if the filesystem supports
locks
or not.
And when the drive is auto mode, use the 'qemu_has_file_lock' to set the toggle.
Signed-off-by: Li Feng
---
v3: don't call the qemu_has_ofd_lock, use a new function instead.
v2: remove the refactoring.
---
block/file-posix.c | 30 +-
include/qemu/osdep.h | 1 +
util/osdep.c | 29 +
3 files changed, 47 insertions(+), 13 deletions(-)
diff --git a/block/file-posix.c b/block/file-posix.c
index 806764f7e3..48f9a32de2 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -606,7 +606,7 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
s->use_lock = false;
break;
case ON_OFF_AUTO_AUTO:
-s->use_lock = qemu_has_ofd_lock();
+s->use_lock = qemu_has_file_lock(filename);
break;
default:
abort();
@@ -2388,6 +2388,7 @@ raw_co_create(BlockdevCreateOptions *options, Error
**errp)
int fd;
uint64_t perm, shared;
int result = 0;
+bool use_lock;
/* Validate options and set default values */
assert(options->driver == BLOCKDEV_DRIVER_FILE);
@@ -2428,19 +2429,22 @@ raw_co_create(BlockdevCreateOptions *options, Error
**errp)
perm = BLK_PERM_WRITE | BLK_PERM_RESIZE;
shared = BLK_PERM_ALL & ~BLK_PERM_RESIZE;
-/* Step one: Take locks */
-result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
-if (result < 0) {
-goto out_close;
-}
+use_lock = qemu_has_file_lock(file_opts->filename);
+if (use_lock) {
+/* Step one: Take locks */
+result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
+if (result < 0) {
+goto out_close;
+}
-/* Step two: Check that nobody else has taken conflicting locks */
-result = raw_check_lock_bytes(fd, perm, shared, errp);
-if (result < 0) {
-error_append_hint(errp,
- "Is another process using the image [%s]?\n",
- file_opts->filename);
-goto out_unlock;
+/* Step two: Check that nobody else has taken conflicting locks */
+result = raw_check_lock_bytes(fd, perm, shared, errp);
+if (result < 0) {
+error_append_hint(errp,
+ "Is another process using the image [%s]?\n",
+ file_opts->filename);
+goto out_unlock;
+}
}
/* Clear the file by truncating it to 0 */
diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h
index f9ec8c84e9..593ae0f4d2 100644
--- a/include/qemu/osdep.h
+++ b/include/qemu/osdep.h
@@ -513,6 +513,7 @@ int qemu_lock_fd(int fd, int64_t start, int64_t len, bool
exclusive);
int qemu_unlock_fd(int fd, int64_t start, int64_t len);
int qemu_lock_fd_test(int fd, int64_t start, int64_t len, bool exclusive);
bool qemu_has_ofd_lock(void);
+bool qemu_has_file_lock(const char *filename);
#endif
#if defined(__HAIKU__) && defined(__i386__)
diff --git a/util/osdep.c b/util/osdep.c
index 66d01b9160..5d9961d261 100644
--- a/util/osdep.c
+++ b/util/osdep.c
@@ -225,6 +225,35 @@ static void qemu_probe_lock_ops(void)
}
}
+bool qemu_has_file_lock(const char *filename)
+{
+#ifdef F_OFD_SETLK
+int cmd = F_OFD_GETLK;
+#else
+int cmd = F_GETLK;
+#endif
+int fd;
+int ret;
+struct flock fl = {
+.l_whence = SEEK_SET,
+.l_start = 0,
+.l_len= 0,
+.l_type = F_WRLCK,
+};
+
+fd = open(filename, O_RDWR);
+if (fd < 0) {
+fprintf(stderr,
+"Failed to open %s for OFD lock probing: %s\n",
+filename,
+strerror(errno));
+return false;
+}
+ret = fcntl(fd, cmd, );
+close(fd);
+return ret == 0;
+}
+
bool qemu_has_ofd_lock(void)
{
qemu_probe_lock_ops();
--
2.24.3
[PATCH v2] file-posix: detect the lock using the real file
This patch addresses this issue:
When accessing a volume on an NFS filesystem without supporting the file lock,
tools, like qemu-img, will complain "Failed to lock byte 100".
In the original code, the qemu_has_ofd_lock will test the lock on the
"/dev/null" pseudo-file. Actually, the file.locking is per-drive property,
which depends on the underlay filesystem.
In this patch, make the 'qemu_has_ofd_lock' with a filename be more
generic and reasonable.
Signed-off-by: Li Feng
---
v2: remove the refactoring.
---
block/file-posix.c | 32 ++--
include/qemu/osdep.h | 2 +-
tests/test-image-locking.c | 2 +-
util/osdep.c | 19 ---
4 files changed, 32 insertions(+), 23 deletions(-)
diff --git a/block/file-posix.c b/block/file-posix.c
index 806764f7e3..03be1b188c 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -595,7 +595,7 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
switch (locking) {
case ON_OFF_AUTO_ON:
s->use_lock = true;
-if (!qemu_has_ofd_lock()) {
+if (!qemu_has_ofd_lock(filename)) {
warn_report("File lock requested but OFD locking syscall is "
"unavailable, falling back to POSIX file locks");
error_printf("Due to the implementation, locks can be lost "
@@ -606,7 +606,7 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
s->use_lock = false;
break;
case ON_OFF_AUTO_AUTO:
-s->use_lock = qemu_has_ofd_lock();
+s->use_lock = qemu_has_ofd_lock(filename);
break;
default:
abort();
@@ -2388,6 +2388,7 @@ raw_co_create(BlockdevCreateOptions *options, Error
**errp)
int fd;
uint64_t perm, shared;
int result = 0;
+bool use_lock;
/* Validate options and set default values */
assert(options->driver == BLOCKDEV_DRIVER_FILE);
@@ -2428,19 +2429,22 @@ raw_co_create(BlockdevCreateOptions *options, Error
**errp)
perm = BLK_PERM_WRITE | BLK_PERM_RESIZE;
shared = BLK_PERM_ALL & ~BLK_PERM_RESIZE;
-/* Step one: Take locks */
-result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
-if (result < 0) {
-goto out_close;
-}
+use_lock = qemu_has_ofd_lock(file_opts->filename);
+if (use_lock) {
+/* Step one: Take locks */
+result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
+if (result < 0) {
+goto out_close;
+}
-/* Step two: Check that nobody else has taken conflicting locks */
-result = raw_check_lock_bytes(fd, perm, shared, errp);
-if (result < 0) {
-error_append_hint(errp,
- "Is another process using the image [%s]?\n",
- file_opts->filename);
-goto out_unlock;
+/* Step two: Check that nobody else has taken conflicting locks */
+result = raw_check_lock_bytes(fd, perm, shared, errp);
+if (result < 0) {
+error_append_hint(errp,
+ "Is another process using the image [%s]?\n",
+ file_opts->filename);
+goto out_unlock;
+}
}
/* Clear the file by truncating it to 0 */
diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h
index f9ec8c84e9..349adad465 100644
--- a/include/qemu/osdep.h
+++ b/include/qemu/osdep.h
@@ -512,7 +512,7 @@ int qemu_dup(int fd);
int qemu_lock_fd(int fd, int64_t start, int64_t len, bool exclusive);
int qemu_unlock_fd(int fd, int64_t start, int64_t len);
int qemu_lock_fd_test(int fd, int64_t start, int64_t len, bool exclusive);
-bool qemu_has_ofd_lock(void);
+bool qemu_has_ofd_lock(const char *filename);
#endif
#if defined(__HAIKU__) && defined(__i386__)
diff --git a/tests/test-image-locking.c b/tests/test-image-locking.c
index ba057bd66c..3e80246081 100644
--- a/tests/test-image-locking.c
+++ b/tests/test-image-locking.c
@@ -149,7 +149,7 @@ int main(int argc, char **argv)
g_test_init(, , NULL);
-if (qemu_has_ofd_lock()) {
+if (qemu_has_ofd_lock(NULL)) {
g_test_add_func("/image-locking/basic", test_image_locking_basic);
g_test_add_func("/image-locking/set-perm-abort", test_set_perm_abort);
}
diff --git a/util/osdep.c b/util/osdep.c
index 66d01b9160..20119aa9ae 100644
--- a/util/osdep.c
+++ b/util/osdep.c
@@ -187,7 +187,7 @@ static int qemu_parse_fdset(const char *param)
return qemu_parse_fd(param);
}
-static void qemu_probe_lock_ops(void)
+static void qemu_probe_lock_ops(const char *filename)
{
if (fcntl_op_setlk == -1) {
#ifdef F_OFD_SETLK
@@ -200,10 +200,15 @@ static void qemu_probe_lock_ops(void)
.l_type = F_WRLCK,
};
-fd = open("/dev/nu
Re: [PATCH] file-posix: detect the lock using the real file
Kevin Wolf 于2020年12月10日周四 上午1:43写道： > > Am 09.12.2020 um 10:33 hat Daniel P. Berrangé geschrieben: > > On Tue, Dec 08, 2020 at 03:38:22PM +0100, Kevin Wolf wrote: > > > Am 08.12.2020 um 13:59 hat Li Feng geschrieben: > > > > This patch addresses this issue: > > > > When accessing a volume on an NFS filesystem without supporting the > > > > file lock, > > > > tools, like qemu-img, will complain "Failed to lock byte 100". > > > > > > > > In the original code, the qemu_has_ofd_lock will test the lock on the > > > > "/dev/null" pseudo-file. Actually, the file.locking is per-drive > > > > property, > > > > which depends on the underlay filesystem. > > > > > > > > In this patch, make the 'qemu_has_ofd_lock' with a filename be more > > > > generic > > > > and reasonable. > > > > > > > > Signed-off-by: Li Feng > > > > > > Do you know any way how I could configure either the NFS server or the > > > NFS client such that locking would fail? For any patch related to this, > > > it would be good if I could even test the scenario. > > > > One could write a qtest that uses an LD_PRELOAD to replace the standard > > glibc fcntl() function with one that returns an error for locking commands. > > Sounds a bit ugly, but for regression testing it could make sense. > > However, part of the testing would be to verify that we our checks > actually match the kernel code, which this approach couldn't solve. > Hi, Kevin and Daniel. How about this patch? I think it's very straightforward. Except we need a mock syscall test case. > Kevin >
[PATCH v2] file-posix: detect the lock using the real file
This patch addresses this issue:
When accessing a volume on an NFS filesystem without supporting the file lock,
tools, like qemu-img, will complain "Failed to lock byte 100".
In the original code, the qemu_has_ofd_lock will test the lock on the
"/dev/null" pseudo-file. Actually, the file.locking is per-drive property,
which depends on the underlay filesystem.
In this patch, make the 'qemu_has_ofd_lock' with a filename be more
generic and reasonable.
Signed-off-by: Li Feng
---
v2: remove the refactoring.
---
block/file-posix.c | 32 ++--
include/qemu/osdep.h | 2 +-
tests/test-image-locking.c | 2 +-
util/osdep.c | 19 ---
4 files changed, 32 insertions(+), 23 deletions(-)
diff --git a/block/file-posix.c b/block/file-posix.c
index 806764f7e3..03be1b188c 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -595,7 +595,7 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
switch (locking) {
case ON_OFF_AUTO_ON:
s->use_lock = true;
-if (!qemu_has_ofd_lock()) {
+if (!qemu_has_ofd_lock(filename)) {
warn_report("File lock requested but OFD locking syscall is "
"unavailable, falling back to POSIX file locks");
error_printf("Due to the implementation, locks can be lost "
@@ -606,7 +606,7 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
s->use_lock = false;
break;
case ON_OFF_AUTO_AUTO:
-s->use_lock = qemu_has_ofd_lock();
+s->use_lock = qemu_has_ofd_lock(filename);
break;
default:
abort();
@@ -2388,6 +2388,7 @@ raw_co_create(BlockdevCreateOptions *options, Error
**errp)
int fd;
uint64_t perm, shared;
int result = 0;
+bool use_lock;
/* Validate options and set default values */
assert(options->driver == BLOCKDEV_DRIVER_FILE);
@@ -2428,19 +2429,22 @@ raw_co_create(BlockdevCreateOptions *options, Error
**errp)
perm = BLK_PERM_WRITE | BLK_PERM_RESIZE;
shared = BLK_PERM_ALL & ~BLK_PERM_RESIZE;
-/* Step one: Take locks */
-result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
-if (result < 0) {
-goto out_close;
-}
+use_lock = qemu_has_ofd_lock(file_opts->filename);
+if (use_lock) {
+/* Step one: Take locks */
+result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
+if (result < 0) {
+goto out_close;
+}
-/* Step two: Check that nobody else has taken conflicting locks */
-result = raw_check_lock_bytes(fd, perm, shared, errp);
-if (result < 0) {
-error_append_hint(errp,
- "Is another process using the image [%s]?\n",
- file_opts->filename);
-goto out_unlock;
+/* Step two: Check that nobody else has taken conflicting locks */
+result = raw_check_lock_bytes(fd, perm, shared, errp);
+if (result < 0) {
+error_append_hint(errp,
+ "Is another process using the image [%s]?\n",
+ file_opts->filename);
+goto out_unlock;
+}
}
/* Clear the file by truncating it to 0 */
diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h
index f9ec8c84e9..349adad465 100644
--- a/include/qemu/osdep.h
+++ b/include/qemu/osdep.h
@@ -512,7 +512,7 @@ int qemu_dup(int fd);
int qemu_lock_fd(int fd, int64_t start, int64_t len, bool exclusive);
int qemu_unlock_fd(int fd, int64_t start, int64_t len);
int qemu_lock_fd_test(int fd, int64_t start, int64_t len, bool exclusive);
-bool qemu_has_ofd_lock(void);
+bool qemu_has_ofd_lock(const char *filename);
#endif
#if defined(__HAIKU__) && defined(__i386__)
diff --git a/tests/test-image-locking.c b/tests/test-image-locking.c
index ba057bd66c..3e80246081 100644
--- a/tests/test-image-locking.c
+++ b/tests/test-image-locking.c
@@ -149,7 +149,7 @@ int main(int argc, char **argv)
g_test_init(, , NULL);
-if (qemu_has_ofd_lock()) {
+if (qemu_has_ofd_lock(NULL)) {
g_test_add_func("/image-locking/basic", test_image_locking_basic);
g_test_add_func("/image-locking/set-perm-abort", test_set_perm_abort);
}
diff --git a/util/osdep.c b/util/osdep.c
index 66d01b9160..20119aa9ae 100644
--- a/util/osdep.c
+++ b/util/osdep.c
@@ -187,7 +187,7 @@ static int qemu_parse_fdset(const char *param)
return qemu_parse_fd(param);
}
-static void qemu_probe_lock_ops(void)
+static void qemu_probe_lock_ops(const char *filename)
{
if (fcntl_op_setlk == -1) {
#ifdef F_OFD_SETLK
@@ -200,10 +200,15 @@ static void qemu_probe_lock_ops(void)
.l_type = F_WRLCK,
};
-fd = open("/dev/nu
Re: [PATCH] file-posix: detect the lock using the real file
Kevin Wolf 于2020年12月8日周二 下午10:38写道： > > Am 08.12.2020 um 13:59 hat Li Feng geschrieben: > > This patch addresses this issue: > > When accessing a volume on an NFS filesystem without supporting the file > > lock, > > tools, like qemu-img, will complain "Failed to lock byte 100". > > > > In the original code, the qemu_has_ofd_lock will test the lock on the > > "/dev/null" pseudo-file. Actually, the file.locking is per-drive property, > > which depends on the underlay filesystem. > > > > In this patch, make the 'qemu_has_ofd_lock' with a filename be more generic > > and reasonable. > > > > Signed-off-by: Li Feng > > Do you know any way how I could configure either the NFS server or the > NFS client such that locking would fail? For any patch related to this, > it would be good if I could even test the scenario. > Hi Kevin, currently our SmartX ZBS storage NFS server doesn't support the file lock and the lock operation will return failure. I have tried the kernel NFS server, and it works well. I don't have more kinds of NFS servers. > For this specific patch, I think Daniel has already provided a good > explanation of the fundamental problems it has. > > Kevin >
Re: [PATCH] file-posix: detect the lock using the real file
Daniel P. Berrangé 于2020年12月8日周二 下午9:45写道：
>
> On Tue, Dec 08, 2020 at 08:59:37PM +0800, Li Feng wrote:
> > This patch addresses this issue:
> > When accessing a volume on an NFS filesystem without supporting the file
> > lock,
> > tools, like qemu-img, will complain "Failed to lock byte 100".
> >
> > In the original code, the qemu_has_ofd_lock will test the lock on the
> > "/dev/null" pseudo-file. Actually, the file.locking is per-drive property,
> > which depends on the underlay filesystem.
>
> IIUC, the problem you're describing is one of whether the filesystem
> supports fcntl locking at all, which is indeed a per-FS check.
>
> The QEMU code being changed though is just about detecting whether
> the host OS supports OFD to not, which is supposed to be a kernel
> level feature applied universally to all FS types.
>
> >
> > In this patch, make the 'qemu_has_ofd_lock' with a filename be more generic
> > and reasonable.
> >
> > Signed-off-by: Li Feng
> > ---
> > block/file-posix.c | 32 +++-
> > include/qemu/osdep.h | 2 +-
> > tests/test-image-locking.c | 2 +-
> > util/osdep.c | 43 --
> > 4 files changed, 47 insertions(+), 32 deletions(-)
> >
> > diff --git a/block/file-posix.c b/block/file-posix.c
> > index 806764f7e3..03be1b188c 100644
> > --- a/block/file-posix.c
> > +++ b/block/file-posix.c
> > @@ -595,7 +595,7 @@ static int raw_open_common(BlockDriverState *bs, QDict
> > *options,
> > switch (locking) {
> > case ON_OFF_AUTO_ON:
> > s->use_lock = true;
> > -if (!qemu_has_ofd_lock()) {
> > +if (!qemu_has_ofd_lock(filename)) {
> > warn_report("File lock requested but OFD locking syscall is "
> > "unavailable, falling back to POSIX file locks");
> > error_printf("Due to the implementation, locks can be lost "
> > @@ -606,7 +606,7 @@ static int raw_open_common(BlockDriverState *bs, QDict
> > *options,
> > s->use_lock = false;
> > break;
> > case ON_OFF_AUTO_AUTO:
> > -s->use_lock = qemu_has_ofd_lock();
> > +s->use_lock = qemu_has_ofd_lock(filename);
> > break;
> > default:
> > abort();
> > @@ -2388,6 +2388,7 @@ raw_co_create(BlockdevCreateOptions *options, Error
> > **errp)
> > int fd;
> > uint64_t perm, shared;
> > int result = 0;
> > +bool use_lock;
> >
> > /* Validate options and set default values */
> > assert(options->driver == BLOCKDEV_DRIVER_FILE);
> > @@ -2428,19 +2429,22 @@ raw_co_create(BlockdevCreateOptions *options, Error
> > **errp)
> > perm = BLK_PERM_WRITE | BLK_PERM_RESIZE;
> > shared = BLK_PERM_ALL & ~BLK_PERM_RESIZE;
> >
> > -/* Step one: Take locks */
> > -result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
> > -if (result < 0) {
> > -goto out_close;
> > -}
> > +use_lock = qemu_has_ofd_lock(file_opts->filename);
> > +if (use_lock) {
> > +/* Step one: Take locks */
> > +result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false,
> > errp);
> > +if (result < 0) {
> > +goto out_close;
> > +}
> >
> > -/* Step two: Check that nobody else has taken conflicting locks */
> > -result = raw_check_lock_bytes(fd, perm, shared, errp);
> > -if (result < 0) {
> > -error_append_hint(errp,
> > - "Is another process using the image [%s]?\n",
> > - file_opts->filename);
> > -goto out_unlock;
> > +/* Step two: Check that nobody else has taken conflicting locks */
> > +result = raw_check_lock_bytes(fd, perm, shared, errp);
> > +if (result < 0) {
> > +error_append_hint(errp,
> > + "Is another process using the image [%s]?\n",
> > + file_opts->filename);
> > +goto out_unlock;
> > +}
> > }
> >
> > /* Clear the file by truncating it to 0 */
> > diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h
> > index f9ec8c84e9..349adad465 100644
> > --- a/include/qemu/osdep.h
> > +++ b/include/qemu/osdep.h
> > @@ -512,7 +5
[PATCH] file-posix: detect the lock using the real file
This patch addresses this issue:
When accessing a volume on an NFS filesystem without supporting the file lock,
tools, like qemu-img, will complain "Failed to lock byte 100".
In the original code, the qemu_has_ofd_lock will test the lock on the
"/dev/null" pseudo-file. Actually, the file.locking is per-drive property,
which depends on the underlay filesystem.
In this patch, make the 'qemu_has_ofd_lock' with a filename be more generic
and reasonable.
Signed-off-by: Li Feng
---
block/file-posix.c | 32 +++-
include/qemu/osdep.h | 2 +-
tests/test-image-locking.c | 2 +-
util/osdep.c | 43 --
4 files changed, 47 insertions(+), 32 deletions(-)
diff --git a/block/file-posix.c b/block/file-posix.c
index 806764f7e3..03be1b188c 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -595,7 +595,7 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
switch (locking) {
case ON_OFF_AUTO_ON:
s->use_lock = true;
-if (!qemu_has_ofd_lock()) {
+if (!qemu_has_ofd_lock(filename)) {
warn_report("File lock requested but OFD locking syscall is "
"unavailable, falling back to POSIX file locks");
error_printf("Due to the implementation, locks can be lost "
@@ -606,7 +606,7 @@ static int raw_open_common(BlockDriverState *bs, QDict
*options,
s->use_lock = false;
break;
case ON_OFF_AUTO_AUTO:
-s->use_lock = qemu_has_ofd_lock();
+s->use_lock = qemu_has_ofd_lock(filename);
break;
default:
abort();
@@ -2388,6 +2388,7 @@ raw_co_create(BlockdevCreateOptions *options, Error
**errp)
int fd;
uint64_t perm, shared;
int result = 0;
+bool use_lock;
/* Validate options and set default values */
assert(options->driver == BLOCKDEV_DRIVER_FILE);
@@ -2428,19 +2429,22 @@ raw_co_create(BlockdevCreateOptions *options, Error
**errp)
perm = BLK_PERM_WRITE | BLK_PERM_RESIZE;
shared = BLK_PERM_ALL & ~BLK_PERM_RESIZE;
-/* Step one: Take locks */
-result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
-if (result < 0) {
-goto out_close;
-}
+use_lock = qemu_has_ofd_lock(file_opts->filename);
+if (use_lock) {
+/* Step one: Take locks */
+result = raw_apply_lock_bytes(NULL, fd, perm, ~shared, false, errp);
+if (result < 0) {
+goto out_close;
+}
-/* Step two: Check that nobody else has taken conflicting locks */
-result = raw_check_lock_bytes(fd, perm, shared, errp);
-if (result < 0) {
-error_append_hint(errp,
- "Is another process using the image [%s]?\n",
- file_opts->filename);
-goto out_unlock;
+/* Step two: Check that nobody else has taken conflicting locks */
+result = raw_check_lock_bytes(fd, perm, shared, errp);
+if (result < 0) {
+error_append_hint(errp,
+ "Is another process using the image [%s]?\n",
+ file_opts->filename);
+goto out_unlock;
+}
}
/* Clear the file by truncating it to 0 */
diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h
index f9ec8c84e9..349adad465 100644
--- a/include/qemu/osdep.h
+++ b/include/qemu/osdep.h
@@ -512,7 +512,7 @@ int qemu_dup(int fd);
int qemu_lock_fd(int fd, int64_t start, int64_t len, bool exclusive);
int qemu_unlock_fd(int fd, int64_t start, int64_t len);
int qemu_lock_fd_test(int fd, int64_t start, int64_t len, bool exclusive);
-bool qemu_has_ofd_lock(void);
+bool qemu_has_ofd_lock(const char *filename);
#endif
#if defined(__HAIKU__) && defined(__i386__)
diff --git a/tests/test-image-locking.c b/tests/test-image-locking.c
index ba057bd66c..3e80246081 100644
--- a/tests/test-image-locking.c
+++ b/tests/test-image-locking.c
@@ -149,7 +149,7 @@ int main(int argc, char **argv)
g_test_init(, , NULL);
-if (qemu_has_ofd_lock()) {
+if (qemu_has_ofd_lock(NULL)) {
g_test_add_func("/image-locking/basic", test_image_locking_basic);
g_test_add_func("/image-locking/set-perm-abort", test_set_perm_abort);
}
diff --git a/util/osdep.c b/util/osdep.c
index 66d01b9160..e7e502edd1 100644
--- a/util/osdep.c
+++ b/util/osdep.c
@@ -42,6 +42,7 @@ extern int madvise(char *, size_t, int);
static bool fips_enabled = false;
static const char *hw_version = QEMU_HW_VERSION;
+static const char *null_device = "/dev/null";
int socket_set_cork(int fd, int v)
{
@@ -187,11 +188,10 @@ static int qemu_parse_fdset(const char *param)
return qemu_parse_fd(param);
}
-static void qemu_probe_lock_ops(void)
+static void qemu_probe_lock_ops_fd(in
Re: [PATCH] file-posix: check the use_lock
Hi Kevin,
OK, thanks for your reply, I will send out the v2 patch.
Thanks,
Feng Li
Kevin Wolf 于2020年12月7日周一 下午7:17写道：
>
> Am 07.12.2020 um 11:50 hat Feng Li geschrieben:
> > Hi Kevin,
> >
> > I have read the 182 test, and I'm very confused about the test.
> > I'm not familiar with the permissions and locks in the qemu.
> > Could you give me more tips about how to complete the test?
>
> Hm, actually, to produce a failure, we would have to have a filesystem
> in the host that doesn't support locks. I don't even know how to get
> such a filesystem manually, and it's probably completely impossible in a
> test case without root permissions.
>
> So maybe just add a more detailed description of the bug to the commit
> message, and we'll have to apply it without a test.
>
> Kevin
>
> > Li Feng 于2020年12月4日周五 下午6:55写道：
> > >
> > > Hi Kevin,
> > > Thanks for your reply.
> > >
> > > In my scenario, my NFS server doesn't support the file lock.
> > > And when I set the file.locking = false, the Qemu still reports:
> > > qemu-system-x86_64: -drive
> > > file=/tmp/nfs/a,format=raw,cache=none,aio=native,if=none,id=drive-virtio-disk1,file.locking=on:
> > > Failed to lock byte 100
> > >
> > > I will look at the iotest 182 and try to add a test.
> > >
> > > Thanks,
> > > Feng Li
> > >
> > > Kevin Wolf 于2020年12月4日周五 下午6:40写道：
> > > >
> > > > Am 04.12.2020 um 11:28 hat Li Feng geschrieben:
> > > > > When setting the file.locking = false, we shouldn't set the lock.
> > > > >
> > > > > Signed-off-by: Li Feng
> > > >
> > > > This looks right to me, but can you add a test for this scenario to
> > > > iotest 182? This would both demonstrate the effect of the bug (I think
> > > > it would be that files are locked after reopen even with locking=off?)
> > > > and make sure that we won't have a regression later. Mentioning the
> > > > effect in the commit message would be good, too.
> > > >
> > > > Kevin
> > > >
> > > > > block/file-posix.c | 2 +-
> > > > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > > > >
> > > > > diff --git a/block/file-posix.c b/block/file-posix.c
> > > > > index d5fd1dbcd2..806764f7e3 100644
> > > > > --- a/block/file-posix.c
> > > > > +++ b/block/file-posix.c
> > > > > @@ -3104,7 +3104,7 @@ static int raw_check_perm(BlockDriverState *bs,
> > > > > uint64_t perm, uint64_t shared,
> > > > > }
> > > > >
> > > > > /* Copy locks to the new fd */
> > > > > -if (s->perm_change_fd) {
> > > > > +if (s->perm_change_fd && s->use_lock) {
> > > > > ret = raw_apply_lock_bytes(NULL, s->perm_change_fd, perm,
> > > > > ~shared,
> > > > > false, errp);
> > > > > if (ret < 0) {
> > > > > --
> > > > > 2.24.3
> > > > >
> > > >
> >
>
[PATCH v2] file-posix: check the use_lock before setting the file lock
The scenario is that when accessing a volume on an NFS filesystem
without supporting the file lock, Qemu will complain "Failed to lock
byte 100", even when setting the file.locking = off.
We should do file lock related operations only when the file.locking is
enabled, otherwise, the syscall of 'fcnctl' will return non-zero.
Signed-off-by: Li Feng
---
block/file-posix.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/block/file-posix.c b/block/file-posix.c
index d5fd1dbcd2..806764f7e3 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -3104,7 +3104,7 @@ static int raw_check_perm(BlockDriverState *bs, uint64_t
perm, uint64_t shared,
}
/* Copy locks to the new fd */
-if (s->perm_change_fd) {
+if (s->perm_change_fd && s->use_lock) {
ret = raw_apply_lock_bytes(NULL, s->perm_change_fd, perm, ~shared,
false, errp);
if (ret < 0) {
--
2.24.3
Re: [PATCH] file-posix: check the use_lock
Hi Kevin,
Thanks for your reply.
In my scenario, my NFS server doesn't support the file lock.
And when I set the file.locking = false, the Qemu still reports:
qemu-system-x86_64: -drive
file=/tmp/nfs/a,format=raw,cache=none,aio=native,if=none,id=drive-virtio-disk1,file.locking=on:
Failed to lock byte 100
I will look at the iotest 182 and try to add a test.
Thanks,
Feng Li
Kevin Wolf 于2020年12月4日周五 下午6:40写道：
>
> Am 04.12.2020 um 11:28 hat Li Feng geschrieben:
> > When setting the file.locking = false, we shouldn't set the lock.
> >
> > Signed-off-by: Li Feng
>
> This looks right to me, but can you add a test for this scenario to
> iotest 182? This would both demonstrate the effect of the bug (I think
> it would be that files are locked after reopen even with locking=off?)
> and make sure that we won't have a regression later. Mentioning the
> effect in the commit message would be good, too.
>
> Kevin
>
> > block/file-posix.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/block/file-posix.c b/block/file-posix.c
> > index d5fd1dbcd2..806764f7e3 100644
> > --- a/block/file-posix.c
> > +++ b/block/file-posix.c
> > @@ -3104,7 +3104,7 @@ static int raw_check_perm(BlockDriverState *bs,
> > uint64_t perm, uint64_t shared,
> > }
> >
> > /* Copy locks to the new fd */
> > -if (s->perm_change_fd) {
> > +if (s->perm_change_fd && s->use_lock) {
> > ret = raw_apply_lock_bytes(NULL, s->perm_change_fd, perm, ~shared,
> > false, errp);
> > if (ret < 0) {
> > --
> > 2.24.3
> >
>
[PATCH] file-posix: check the use_lock
When setting the file.locking = false, we shouldn't set the lock.
Signed-off-by: Li Feng
---
block/file-posix.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/block/file-posix.c b/block/file-posix.c
index d5fd1dbcd2..806764f7e3 100644
--- a/block/file-posix.c
+++ b/block/file-posix.c
@@ -3104,7 +3104,7 @@ static int raw_check_perm(BlockDriverState *bs, uint64_t
perm, uint64_t shared,
}
/* Copy locks to the new fd */
-if (s->perm_change_fd) {
+if (s->perm_change_fd && s->use_lock) {
ret = raw_apply_lock_bytes(NULL, s->perm_change_fd, perm, ~shared,
false, errp);
if (ret < 0) {
--
2.24.3
Re: [PATCH v4 2/2] vhost-user-blk: delay vhost_user_blk_disconnect
Hi Raphael,
I'm sorry. I just end my journey today.
Yes, pls sign off me here.
this patch is nearly the same as my previous patch.
Thanks,
Feng Li
Raphael Norwitz 于2020年5月31日周日 上午8:55写道：
>
> On Thu, May 28, 2020 at 5:13 AM Dima Stepanov wrote:
> >
> > A socket write during vhost-user communication may trigger a disconnect
> > event, calling vhost_user_blk_disconnect() and clearing all the
> > vhost_dev structures holding data that vhost-user functions expect to
> > remain valid to roll back initialization correctly. Delay the cleanup to
> > keep vhost_dev structure valid.
> > There are two possible states to handle:
> > 1. RUN_STATE_PRELAUNCH: skip bh oneshot call and perform disconnect in
> > the caller routine.
> > 2. RUN_STATE_RUNNING: delay by using bh
> >
> > BH changes are based on the similar changes for the vhost-user-net
> > device:
> > commit e7c83a885f865128ae3cf1946f8cb538b63cbfba
> > "vhost-user: delay vhost_user_stop"
> >
> > Signed-off-by: Dima Stepanov
>
> Reviewed-by: Raphael Norwitz
>
> Li Feng - would you also like to sign off here?
>
> > ---
> > hw/block/vhost-user-blk.c | 38 +-
> > 1 file changed, 37 insertions(+), 1 deletion(-)
> >
> > diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
> > index 9d8c0b3..76838e7 100644
> > --- a/hw/block/vhost-user-blk.c
> > +++ b/hw/block/vhost-user-blk.c
> > @@ -349,6 +349,19 @@ static void vhost_user_blk_disconnect(DeviceState *dev)
> > vhost_dev_cleanup(>dev);
> > }
> >
> > +static void vhost_user_blk_event(void *opaque, QEMUChrEvent event);
> > +
> > +static void vhost_user_blk_chr_closed_bh(void *opaque)
> > +{
> > +DeviceState *dev = opaque;
> > +VirtIODevice *vdev = VIRTIO_DEVICE(dev);
> > +VHostUserBlk *s = VHOST_USER_BLK(vdev);
> > +
> > +vhost_user_blk_disconnect(dev);
> > +qemu_chr_fe_set_handlers(>chardev, NULL, NULL, vhost_user_blk_event,
> > +NULL, opaque, NULL, true);
> > +}
> > +
> > static void vhost_user_blk_event(void *opaque, QEMUChrEvent event)
> > {
> > DeviceState *dev = opaque;
> > @@ -363,7 +376,30 @@ static void vhost_user_blk_event(void *opaque,
> > QEMUChrEvent event)
> > }
> > break;
> > case CHR_EVENT_CLOSED:
> > -vhost_user_blk_disconnect(dev);
> > +/*
> > + * A close event may happen during a read/write, but vhost
> > + * code assumes the vhost_dev remains setup, so delay the
> > + * stop & clear. There are two possible paths to hit this
> > + * disconnect event:
> > + * 1. When VM is in the RUN_STATE_PRELAUNCH state. The
> > + * vhost_user_blk_device_realize() is a caller.
> > + * 2. In tha main loop phase after VM start.
> > + *
> > + * For p2 the disconnect event will be delayed. We can't
> > + * do the same for p1, because we are not running the loop
> > + * at this moment. So just skip this step and perform
> > + * disconnect in the caller function.
> > + *
> > + * TODO: maybe it is a good idea to make the same fix
> > + * for other vhost-user devices.
> > + */
> > +if (runstate_is_running()) {
> > +AioContext *ctx = qemu_get_current_aio_context();
> > +
> > +qemu_chr_fe_set_handlers(>chardev, NULL, NULL, NULL, NULL,
> > +NULL, NULL, false);
> > +aio_bh_schedule_oneshot(ctx, vhost_user_blk_chr_closed_bh,
> > opaque);
> > +}
> > break;
> > case CHR_EVENT_BREAK:
> > case CHR_EVENT_MUX_IN:
> > --
> > 2.7.4
> >
> >
Re: [PATCH v2 5/5] vhost: add device started check in migration set log
Hi, Dima.
This abort is what I have mentioned in my previous email.
I have triggered this crash without any fix a week ago.
And I have written a test patch to let vhost_log_global_start return
int and propagate the error to up layer.
However, my change is a little large, because the origin callback
return void, and don't do some rollback.
After test, the migration could migrate to dst successfully, and fio
is still running perfectly, but the src vm is still stuck here, no
crash.
Is it right to return this error to the up layer?
Thanks,
Feng Li
Dima Stepanov 于2020年5月16日周六 上午12:55写道：
>
> On Thu, May 14, 2020 at 03:34:24PM +0800, Jason Wang wrote:
> >
> > On 2020/5/13 下午5:47, Dima Stepanov wrote:
> > >>> case CHR_EVENT_CLOSED:
> > >>> /* a close event may happen during a read/write, but vhost
> > >>> * code assumes the vhost_dev remains setup, so delay the
> > >>> * stop & clear to idle.
> > >>> * FIXME: better handle failure in vhost code, remove bh
> > >>> */
> > >>> if (s->watch) {
> > >>> AioContext *ctx = qemu_get_current_aio_context();
> > >>>
> > >>> g_source_remove(s->watch);
> > >>> s->watch = 0;
> > >>> qemu_chr_fe_set_handlers(>chr, NULL, NULL, NULL, NULL,
> > >>> NULL, NULL, false);
> > >>>
> > >>> aio_bh_schedule_oneshot(ctx, chr_closed_bh, opaque);
> > >>> }
> > >>> break;
> > >>>
> > >>>I think it's time we dropped the FIXME and moved the handling to common
> > >>>code. Jason? Marc-André?
> > >>I agree. Just to confirm, do you prefer bh or doing changes like what is
> > >>done in this series? It looks to me bh can have more easier codes.
> > >Could it be a good idea just to make disconnect in the char device but
> > >postphone clean up in the vhost-user-blk (or any other vhost-user
> > >device) itself? So we are moving the postphone logic and decision from
> > >the char device to vhost-user device. One of the idea i have is as
> > >follows:
> > > - Put ourself in the INITIALIZATION state
> > > - Start these vhost-user "handshake" commands
> > > - If we got a disconnect error, perform disconnect, but don't clean up
> > > device (it will be clean up on the roll back). I can be done by
> > > checking the state in vhost_user_..._disconnect routine or smth like
> > > it
> >
> >
> > Any issue you saw just using the aio bh as Michael posted above.
> >
> > Then we don't need to deal with the silent vhost_dev_stop() and we will have
> > codes that is much more easier to understand.
> I've implemented this solution inside
> hw/block/vhost-user-blk.c:vhost_user_blk_event() in the similar way by
> using the s->connected field. Looks good and more correct fix ). I have
> two questions here before i'll rework the fixes:
> 1. Is it okay to make the similar fix inside vhost_user_blk_event() or
> we are looking for more generic vhost-user solution? What do you think?
> 2. For migration we require an additional information that for the
> vhost-user device it isn't an error, because i'm trigerring the
> following assert error:
> Core was generated by `x86_64-softmmu/qemu-system-x86_64 -nodefaults
> -no-user-config -M q35,sata=false'.
> Program terminated with signal SIGABRT, Aborted.
> #0 0x7fb56e729428 in raise () from /lib/x86_64-linux-gnu/libc.so.6
> [Current thread is 1 (Thread 0x7fb486ef5700 (LWP 527734))]
>
> (gdb) bt
> #0 0x7fb56e729428 in raise () from /lib/x86_64-linux-gnu/libc.so.6
> #1 0x7fb56e72b02a in abort () from /lib/x86_64-linux-gnu/libc.so.6
> #2 0x5648ea376ee6 in vhost_log_global_start
> (listener=0x5648ece4eb08) at ./hw/virtio/vhost.c:857
> #3 0x5648ea2dde7e in memory_global_dirty_log_start ()
> at ./memory.c:2611
> #4 0x5648ea2e68e7 in ram_init_bitmaps (rs=0x7fb4740008c0)
> at ./migration/ram.c:2305
> #5 0x5648ea2e698b in ram_init_all (rsp=0x5648eb1f0f20 )
> at ./migration/ram.c:2323
> #6 0x5648ea2e6cc5 in ram_save_setup (f=0x5648ec609e00,
> opaque=0x5648eb1f0f20 )
> at ./migration/ram.c:2436
> #7 0x5648ea67b7d3 in qemu_savevm_state_setup (f=0x5648ec609e00) at
> migration/savevm.c:1176
> #8 0x5648ea674511 in migration_thread (opaque=0x5648ec031ff0) at
> migration/migration.c:3416
> #9 0x5648ea85d65d in qemu_thread_start (args=0x5648ec6057f0) at
> util/qemu-thread-posix.c:519
> #10 0x7fb56eac56ba in start_thread () from
> /lib/x86_64-linux-gnu/libpthread.so.0
> #11 0x7fb56e7fb41d in clone () from /lib/x86_64-linux-gnu/libc.so.6
> (gdb) frame 2
> #2 0x5648ea376ee6 in vhost_log_global_start
> (listener=0x5648ece4eb08) at ./hw/virtio/vhost.c:857
> 857 abort();
> (gdb) list
> 852 {
> 853 int r;
> 854
> 855 r = vhost_migration_log(listener, true);
> 856 if (r < 0) {
> 857
Re: [PATCH v2 5/5] vhost: add device started check in migration set log
Hi, Dima.
If vhost_migration_log return < 0, then vhost_log_global_start will
trigger a crash.
Does your patch have process this abort?
If a disconnect happens in the migration stage, the correct operation
is to stop the migration, right?
841 static void vhost_log_global_start(MemoryListener *listener)
842 {
843 int r;
844
845 r = vhost_migration_log(listener, true);
846 if (r < 0) {
847 abort();
848 }
849 }
Thanks,
Feng Li
Jason Wang 于2020年5月12日周二 上午11:33写道：
>
>
> On 2020/5/11 下午5:25, Dima Stepanov wrote:
> > On Mon, May 11, 2020 at 11:15:53AM +0800, Jason Wang wrote:
> >> On 2020/4/30 下午9:36, Dima Stepanov wrote:
> >>> If vhost-user daemon is used as a backend for the vhost device, then we
> >>> should consider a possibility of disconnect at any moment. If such
> >>> disconnect happened in the vhost_migration_log() routine the vhost
> >>> device structure will be clean up.
> >>> At the start of the vhost_migration_log() function there is a check:
> >>>if (!dev->started) {
> >>>dev->log_enabled = enable;
> >>>return 0;
> >>>}
> >>> To be consistent with this check add the same check after calling the
> >>> vhost_dev_set_log() routine. This in general help not to break a
> >>> migration due the assert() message. But it looks like that this code
> >>> should be revised to handle these errors more carefully.
> >>>
> >>> In case of vhost-user device backend the fail paths should consider the
> >>> state of the device. In this case we should skip some function calls
> >>> during rollback on the error paths, so not to get the NULL dereference
> >>> errors.
> >>>
> >>> Signed-off-by: Dima Stepanov
> >>> ---
> >>> hw/virtio/vhost.c | 39 +++
> >>> 1 file changed, 35 insertions(+), 4 deletions(-)
> >>>
> >>> diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c
> >>> index 3ee50c4..d5ab96d 100644
> >>> --- a/hw/virtio/vhost.c
> >>> +++ b/hw/virtio/vhost.c
> >>> @@ -787,6 +787,17 @@ static int vhost_dev_set_features(struct vhost_dev
> >>> *dev,
> >>> static int vhost_dev_set_log(struct vhost_dev *dev, bool enable_log)
> >>> {
> >>> int r, i, idx;
> >>> +
> >>> +if (!dev->started) {
> >>> +/*
> >>> + * If vhost-user daemon is used as a backend for the
> >>> + * device and the connection is broken, then the vhost_dev
> >>> + * structure will be reset all its values to 0.
> >>> + * Add additional check for the device state.
> >>> + */
> >>> +return -1;
> >>> +}
> >>> +
> >>> r = vhost_dev_set_features(dev, enable_log);
> >>> if (r < 0) {
> >>> goto err_features;
> >>> @@ -801,12 +812,19 @@ static int vhost_dev_set_log(struct vhost_dev *dev,
> >>> bool enable_log)
> >>> }
> >>> return 0;
> >>> err_vq:
> >>> -for (; i >= 0; --i) {
> >>> +/*
> >>> + * Disconnect with the vhost-user daemon can lead to the
> >>> + * vhost_dev_cleanup() call which will clean up vhost_dev
> >>> + * structure.
> >>> + */
> >>> +for (; dev->started && (i >= 0); --i) {
> >>> idx = dev->vhost_ops->vhost_get_vq_index(
> >>
> >> Why need the check of dev->started here, can started be modified outside
> >> mainloop? If yes, I don't get the check of !dev->started in the beginning
> >> of
> >> this function.
> >>
> > No dev->started can't change outside the mainloop. The main problem is
> > only for the vhost_user_blk daemon. Consider the case when we
> > successfully pass the dev->started check at the beginning of the
> > function, but after it we hit the disconnect on the next call on the
> > second or third iteration:
> > r = vhost_virtqueue_set_addr(dev, dev->vqs + i, idx, enable_log);
> > The unix socket backend device will call the disconnect routine for this
> > device and reset the structure. So the structure will be reset (and
> > dev->started set to false) inside this set_addr() call.
>
>
> I still don't get here. I think the disconnect can not happen in the
> middle of vhost_dev_set_log() since both of them were running in
> mainloop. And even if it can, we probably need other synchronization
> mechanism other than simple check here.
>
>
> > So
> > we shouldn't call the clean up calls because this virtqueues were clean
> > up in the disconnect call. But we should protect these calls somehow, so
> > it will not hit SIGSEGV and we will be able to pass migration.
> >
> > Just to summarize it:
> > For the vhost-user-blk devices we ca hit clean up calls twice in case of
> > vhost disconnect:
> > 1. The first time during the disconnect process. The clean up is called
> > inside it.
> > 2. The second time during roll back clean up.
> > So if it is the case we should skip p2.
> >
> >>> dev, dev->vq_index + i);
> >>> vhost_virtqueue_set_addr(dev, dev->vqs + i, idx,
> >>>dev->log_enabled);
> >>> }
> >>> -vhost_dev_set_features(dev,
Re: [PATCH v2 1/5] char-socket: return -1 in case of disconnect during tcp_chr_write
Thanks,
Feng Li
Dima Stepanov 于2020年4月30日周四 下午9:36写道：
>
> During testing of the vhost-user-blk reconnect functionality the qemu
> SIGSEGV was triggered:
> start qemu as:
> x86_64-softmmu/qemu-system-x86_64 -m 1024M -M q35 \
>-object
> memory-backend-file,id=ram-node0,size=1024M,mem-path=/dev/shm/qemu,share=on \
>-numa node,cpus=0,memdev=ram-node0 \
>-chardev socket,id=chardev0,path=./vhost.sock,noserver,reconnect=1 \
>-device vhost-user-blk-pci,chardev=chardev0,num-queues=4 --enable-kvm
> start vhost-user-blk daemon:
> ./vhost-user-blk -s ./vhost.sock -b test-img.raw
>
> If vhost-user-blk will be killed during the vhost initialization
> process, for instance after getting VHOST_SET_VRING_CALL command, then
> QEMU will fail with the following backtrace:
>
> Thread 1 "qemu-system-x86" received signal SIGSEGV, Segmentation fault.
> 0x559272bb in vhost_user_read (dev=0x7fffef2d53e0, msg=0x7fffd5b0)
> at ./hw/virtio/vhost-user.c:260
> 260 CharBackend *chr = u->user->chr;
>
> #0 0x559272bb in vhost_user_read (dev=0x7fffef2d53e0,
> msg=0x7fffd5b0)
> at ./hw/virtio/vhost-user.c:260
> #1 0x5592acb8 in vhost_user_get_config (dev=0x7fffef2d53e0,
> config=0x7fffef2d5394 "", config_len=60)
> at ./hw/virtio/vhost-user.c:1645
> #2 0x55925525 in vhost_dev_get_config (hdev=0x7fffef2d53e0,
> config=0x7fffef2d5394 "", config_len=60)
> at ./hw/virtio/vhost.c:1490
> #3 0x558cc46b in vhost_user_blk_device_realize (dev=0x7fffef2d51a0,
> errp=0x7fffd8f0)
> at ./hw/block/vhost-user-blk.c:429
> #4 0x55920090 in virtio_device_realize (dev=0x7fffef2d51a0,
> errp=0x7fffd948)
> at ./hw/virtio/virtio.c:3615
> #5 0x55a9779c in device_set_realized (obj=0x7fffef2d51a0,
> value=true, errp=0x7fffdb88)
> at ./hw/core/qdev.c:891
> ...
>
> The problem is that vhost_user_write doesn't get an error after
> disconnect and try to call vhost_user_read(). The tcp_chr_write()
> routine should return -1 in case of disconnect. Indicate the EIO error
> if this routine is called in the disconnected state.
>
> Signed-off-by: Dima Stepanov
> ---
> chardev/char-socket.c | 8 +---
> 1 file changed, 5 insertions(+), 3 deletions(-)
>
> diff --git a/chardev/char-socket.c b/chardev/char-socket.c
> index 185fe38..c128cca 100644
> --- a/chardev/char-socket.c
> +++ b/chardev/char-socket.c
> @@ -175,14 +175,16 @@ static int tcp_chr_write(Chardev *chr, const uint8_t
> *buf, int len)
> if (ret < 0 && errno != EAGAIN) {
> if (tcp_chr_read_poll(chr) <= 0) {
> tcp_chr_disconnect_locked(chr);
> -return len;
> +/* Return an error since we made a disconnect. */
> +return ret;
The `return` statement could be deleted.
The outside has a return statement.
> } /* else let the read handler finish it properly */
> }
>
> return ret;
> } else {
> -/* XXX: indicate an error ? */
> -return len;
> +/* Indicate an error. */
> +errno = EIO;
> +return -1;
> }
> }
>
> --
> 2.7.4
>
Re: [RFC PATCH v1 3/7] char-socket: initialize reconnect timer only if close is emitted
This patch is trying to fix the same issue with me.
However, our fix is different.
I think that check the s->reconnect_timer is better.
Thanks,
Feng Li
Marc-André Lureau 于2020年4月24日周五 上午3:16写道：
>
> Hi
>
> On Thu, Apr 23, 2020 at 8:41 PM Dima Stepanov wrote:
> >
> > During vhost-user reconnect functionality testing the following assert
> > was hit:
> > qemu-system-x86_64: chardev/char-socket.c:125:
> > qemu_chr_socket_restart_timer: Assertion `!s->reconnect_timer' failed.
> > Aborted (core dumped)
>
> That looks related to "[PATCH 3/4] char-socket: avoid double call
> tcp_chr_free_connection"
>
> > This is observed only if the connection is closed by the vhost-user-blk
> > daemon during the initialization routine. In this case the
> > tcp_chr_disconnect_locked() routine is called twice. First time it is
> > called in the tcp_chr_write() routine, after getting the SIGPIPE signal.
> > Second time it is called when vhost_user_blk_connect() routine return
> > error. In general it looks correct, because the initialization routine
> > can return error in many cases.
> > The tcp_chr_disconnect_locked() routine could be fixed. The timer will
> > be restarted only if the close event is emitted.
> >
> > Signed-off-by: Dima Stepanov
> > ---
> > chardev/char-socket.c | 10 +-
> > 1 file changed, 5 insertions(+), 5 deletions(-)
> >
> > diff --git a/chardev/char-socket.c b/chardev/char-socket.c
> > index c128cca..83ca4d9 100644
> > --- a/chardev/char-socket.c
> > +++ b/chardev/char-socket.c
> > @@ -476,7 +476,7 @@ static void update_disconnected_filename(SocketChardev
> > *s)
> > static void tcp_chr_disconnect_locked(Chardev *chr)
> > {
> > SocketChardev *s = SOCKET_CHARDEV(chr);
> > -bool emit_close = s->state == TCP_CHARDEV_STATE_CONNECTED;
> > +bool was_connected = s->state == TCP_CHARDEV_STATE_CONNECTED;
> >
> > tcp_chr_free_connection(chr);
> >
> > @@ -485,11 +485,11 @@ static void tcp_chr_disconnect_locked(Chardev *chr)
> >chr, NULL, chr->gcontext);
> > }
> > update_disconnected_filename(s);
> > -if (emit_close) {
> > +if (was_connected) {
> > qemu_chr_be_event(chr, CHR_EVENT_CLOSED);
> > -}
> > -if (s->reconnect_time) {
> > -qemu_chr_socket_restart_timer(chr);
> > +if (s->reconnect_time) {
> > +qemu_chr_socket_restart_timer(chr);
> > +}
> > }
> > }
> >
> > --
> > 2.7.4
> >
> >
>
>
> --
> Marc-André Lureau
--
The SmartX email address is only for business purpose. Any sent message
that is not related to the business is not authorized or permitted by
SmartX.
本邮箱为北京志凌海纳科技有限公司（SmartX）工作邮箱. 如本邮箱发出的邮件与工作无关,该邮件未得到本公司任何的明示或默示的授权.
Re: [PATCH] vhost-user-blk: fix invalid memory access
Nothing changed. Just separate this patch from those patchset series. Thanks, Feng Li Philippe Mathieu-Daudé 于2020年4月17日周五 下午6:55写道： > > Hi Li, > > On 4/17/20 12:17 PM, Li Feng wrote: > > when s->inflight is freed, vhost_dev_free_inflight may try to access > > s->inflight->addr, it will retrigger the following issue. > > > > ==7309==ERROR: AddressSanitizer: heap-use-after-free on address > > 0x604001020d18 at pc 0x55ce948a bp 0x7fffb170 sp 0x7fffb160 > > READ of size 8 at 0x604001020d18 thread T0 > > #0 0x55ce9489 in vhost_dev_free_inflight > > /root/smartx/qemu-el7/qemu-test/hw/virtio/vhost.c:1473 > > #1 0x55cd86eb in virtio_reset > > /root/smartx/qemu-el7/qemu-test/hw/virtio/virtio.c:1214 > > #2 0x560d3eff in virtio_pci_reset hw/virtio/virtio-pci.c:1859 > > #3 0x55f2ac53 in device_set_realized hw/core/qdev.c:893 > > #4 0x561d572c in property_set_bool qom/object.c:1925 > > #5 0x561de8de in object_property_set_qobject qom/qom-qobject.c:27 > > #6 0x561d99f4 in object_property_set_bool qom/object.c:1188 > > #7 0x55e50ae7 in qdev_device_add > > /root/smartx/qemu-el7/qemu-test/qdev-monitor.c:626 > > Maybe cut <-- > > > #8 0x55e51213 in qmp_device_add > > /root/smartx/qemu-el7/qemu-test/qdev-monitor.c:806 > > #9 0x55e8ff40 in hmp_device_add > > /root/smartx/qemu-el7/qemu-test/hmp.c:1951 > > #10 0x55be889a in handle_hmp_command > > /root/smartx/qemu-el7/qemu-test/monitor.c:3404 > > #11 0x55beac8b in monitor_command_cb > > /root/smartx/qemu-el7/qemu-test/monitor.c:4296 > > #12 0x56433eb7 in readline_handle_byte util/readline.c:393 > > #13 0x55be89ec in monitor_read > > /root/smartx/qemu-el7/qemu-test/monitor.c:4279 > > #14 0x563285cc in tcp_chr_read chardev/char-socket.c:470 > > #15 0x7670b968 in g_main_context_dispatch > > (/lib64/libglib-2.0.so.0+0x4a968) > > #16 0x5640727c in glib_pollfds_poll util/main-loop.c:215 > > #17 0x5640727c in os_host_main_loop_wait util/main-loop.c:238 > > #18 0x5640727c in main_loop_wait util/main-loop.c:497 > > #19 0x55b2d0bf in main_loop > > /root/smartx/qemu-el7/qemu-test/vl.c:2013 > > #20 0x55b2d0bf in main /root/smartx/qemu-el7/qemu-test/vl.c:4776 > > #21 0x7fffdd2eb444 in __libc_start_main (/lib64/libc.so.6+0x22444) > > #22 0x55b3767a > > (/root/smartx/qemu-el7/qemu-test/x86_64-softmmu/qemu-system-x86_64+0x5e367a) > > -->. > > > > > 0x604001020d18 is located 8 bytes inside of 40-byte region > > [0x604001020d10,0x604001020d38) > > freed by thread T0 here: > > #0 0x76f00508 in __interceptor_free (/lib64/libasan.so.4+0xde508) > > #1 0x7671107d in g_free (/lib64/libglib-2.0.so.0+0x5007d) > > > > previously allocated by thread T0 here: > > #0 0x76f00a88 in __interceptor_calloc (/lib64/libasan.so.4+0xdea88) > > #1 0x76710fc5 in g_malloc0 (/lib64/libglib-2.0.so.0+0x4ffc5) > > > > SUMMARY: AddressSanitizer: heap-use-after-free > > /root/smartx/qemu-el7/qemu-test/hw/virtio/vhost.c:1473 in > > vhost_dev_free_inflight > > Shadow bytes around the buggy address: > >0x0c08801fc150: fa fa 00 00 00 00 04 fa fa fa fd fd fd fd fd fa > >0x0c08801fc160: fa fa fd fd fd fd fd fd fa fa 00 00 00 00 04 fa > >0x0c08801fc170: fa fa 00 00 00 00 00 01 fa fa 00 00 00 00 04 fa > >0x0c08801fc180: fa fa 00 00 00 00 00 01 fa fa 00 00 00 00 00 01 > >0x0c08801fc190: fa fa 00 00 00 00 00 fa fa fa 00 00 00 00 04 fa > > =>0x0c08801fc1a0: fa fa fd[fd]fd fd fd fa fa fa fd fd fd fd fd fa > >0x0c08801fc1b0: fa fa fd fd fd fd fd fa fa fa fd fd fd fd fd fa > >0x0c08801fc1c0: fa fa 00 00 00 00 00 fa fa fa fd fd fd fd fd fd > >0x0c08801fc1d0: fa fa 00 00 00 00 00 01 fa fa fd fd fd fd fd fa > >0x0c08801fc1e0: fa fa fd fd fd fd fd fa fa fa fd fd fd fd fd fd > >0x0c08801fc1f0: fa fa 00 00 00 00 00 01 fa fa fd fd fd fd fd fa > > Shadow byte legend (one shadow byte represents 8 application bytes): > >Addressable: 00 > >Partially addressable: 01 02 03 04 05 06 07 > >Heap left redzone: fa > >Freed heap region: fd > > cut <-- > > >Stack left redzone: f1 > >Stack mid redzone: f2 > >Stack right redzone: f3 > >Stack after return: f5 > >Stack use after scope: f8 > >Global redzone: f9 > >Global init order: f6
