Sebastien Boeuf
Sebastien Boeuf
@carezkh thanks for detailing how to reproduce the issue. I'm wondering if this could be related to the issue https://github.com/cloud-hypervisor/cloud-hypervisor/issues/4805 Could you please try the patch proposed by @NunoDasNeves here:...
>@sboeuf I have tested the patch, and it seemed to have resolved the problem, but i'm still not sure if they are related. Thanks for doing that, that might confirm...
@weltling could you please have a look at this issue as it's failing quite often on aarch64.
> @sboeuf thanks for tagging me on this and for preliminary taking care of the sporadic issue. At the time I went for enabling more tests for `aarch64`, I observed...
Now you have the unit tests failing.
What's the status on this?
Starting with Linux v6.0, a new vhost ioctl `VHOST_VDPA_SUSPEND` has been added. It allows the consumer (the VMM) to inform the vDPA device to stop processing the virtqueues, which is...
Hi @eugpermar, I wanted to confirm (after reading the code from the kernel) if the only way to resume a "suspended" vDPA device was by triggering a reset (which means...
>While it's true that the only way to resume a vdpa device is to reset it for the moment it doesn't mean the guest is aware of this reset. >...
>Since the controversial point of block is how to get the in flight buffers, I'd say that a resume operation / ioctl would fit better. That could enable other devices...