Re: [Qemu-devel] [RFC 2/4] hw/i386: Introduce AMD IOMMU

[Valentine Sinitsyn]

On 25.08.2015 12:41, David kiarie wrote:
> On Tue, Aug 25, 2015 at 10:31 AM, Valentine Sinitsyn
[...snip...]
> > The spec doesn't say anything about this, I checked it. IIRC, I came across
> > this behavior when submitted COMPLETION_WAIT command with interrupts off and
> > spin in a while() loop waiting for the mask to be written. The loop never
> > finished, but everything worked as intended if I removed COM_COMPLETION_INTR
> > for the command.
>
> Is this behaviour useful in any way ?I'll try to fix this in the next version.
I don't think it's useful. The reason I posted this question was to 
check if this is a bug in silicon or intended behavior. If the latter, 
I'd suggest to emulate it.

Valentine