Re: [Xen-devel] [PATCH 00/25 v7] SBSA UART emulation support in Xen

[Julien Grall <julien.grall@xxxxxxx>]

Hi Wei,

On 15/08/17 10:49, Wei Liu wrote:
> On Thu, Aug 10, 2017 at 06:51:24PM +0100, Julien Grall wrote:
> > > > > The interface between Xen and xenconsoled can be asynchronous, it can
> > > > > opt to queue X characters before sending an event, also setup a oneshot
> > > > > timer to avoid hanging.
> > > > >
> > > > > This however has some other implications -- it might not be as reliable
> > > > > as the original method because data is not guaranteed to hit backend. If
> > > > > the guest crashes very early on, depending the actual implementation you
> > > > > might not be able get the data.
> > > > Would it be possible to ask xenconsoled to dump everything on domain crash?
> > > > Some kind of synchronization.
> > > No, not at the moment. If the data is still in Xen and destroyed,
> > > xenconsoled can't do anything.
> > The vUART emulation is directly queuing the data, there are no intermediate
> > buffer. So all the data would be in the shared ring available for
> > xenconsoled to go through.
> >
> > It would be quite a useful enhancement for when the guest crash.
> What I meant was actually "If the data is still in the ring". There is
> no synchronisation between Xen and xenconsoled to let the latter pull
> out all data before destroying the guest.
I don't think you would need synchronisation between Xen and xenconsoled 
at domain destruction. Given that xenconsoled would have to unmap the 
page, I was suggesting that it makes sure that cons == prod before 
destroying the instance. So all the character queued would be displayed 
on the console.
> As it stands, if we go with the fully-synchronised approach for now,
> that won't be a problem. But when you want to fiddle with the BUSY bit
> or any other optimisation -- leaving more data in the ring -- that might
> cause data loss.
The fully-synchronised solution is far too slow, we are looking at 
optimization to avoid the synchronization for every single character. 
This would leave some potential data loss (max a few characters) if the 
domain crashed.
> I'm not against bumping the rate-limit, but I want a justification and
> want to consider as many approaches as possible. Ultimately the final
> decision is up to you and Stefano.
I am not in favor of bumping the rate-limit at the moment. I believe we 
can solve the problem directly in the emulation. I spoke with Bhupinder 
about various solution that seem promising.
Cheers,

--
Julien Grall

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
https://lists.xen.org/xen-devel