[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH] xc_save: ignore the first suspend event channel notification


  • To: Brendan Cully <brendan@xxxxxxxxx>
  • From: Keir Fraser <keir.fraser@xxxxxxxxxxxxx>
  • Date: Sun, 07 Sep 2008 08:43:24 +0100
  • Cc: xen-devel@xxxxxxxxxxxxxxxxxxx
  • Delivery-date: Sun, 07 Sep 2008 00:43:27 -0700
  • List-id: Xen developer discussion <xen-devel.lists.xensource.com>
  • Thread-index: AckQvWg1poLqMnywEd2cRgAWy6hiGQ==
  • Thread-topic: [Xen-devel] [PATCH] xc_save: ignore the first suspend event channel notification

On 7/9/08 08:15, "Brendan Cully" <brendan@xxxxxxxxx> wrote:

>>> I don't like the forced initial consumption of a single evtchn signal. I'd
>>> like rid of that and change the retry loop to call (*suspend) instead. Add
>>> an extra boolean to the static suspend_info structure to detect first
>>> invocation of (*suspend) versus repeated invocations.
>> 
>> Actually I can knock these changes together myself, so I'll do that, taking
>> whatever bits of your patch are appropriate.
> 
> Ok, sure. I still think it's a bit unfortunate that the suspend
> function will return without waiting for the domain to suspend, and
> not just because of the extra round trip. At the end of the suspend
> function, xend is under the impression that the guest has suspended
> and it can perform device migration stage 2. I wonder if it's
> dangerous to do this before confirming that the guest has actually
> suspended.

Ah, I hadn't spotted that bit. I'll think about it - maybe your approach is
simplest after all.

 -- Keir



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.