This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
Home Products Support Community News


Re: [GIT/PATCH 0/5] Re: [Xen-devel] unfair servicing of DomU vbd request

To: John Weekes <lists.xen@xxxxxxxxxxxxxxxxxx>
Subject: Re: [GIT/PATCH 0/5] Re: [Xen-devel] unfair servicing of DomU vbd requests
From: Ian Campbell <Ian.Campbell@xxxxxxxxxxxxx>
Date: Fri, 4 Mar 2011 09:15:27 +0000
Cc: "xen-devel@xxxxxxxxxxxxxxxxxxx" <xen-devel@xxxxxxxxxxxxxxxxxxx>
Delivery-date: Fri, 04 Mar 2011 01:16:22 -0800
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
In-reply-to: <4D70A59A.2060603@xxxxxxxxxxxxxxxxxx>
List-help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-id: Xen developer discussion <xen-devel.lists.xensource.com>
List-post: <mailto:xen-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
Organization: Citrix Systems, Inc.
References: <C995020F.1412D%keir.xen@xxxxxxxxx> <1299172198.6552.14.camel@xxxxxxxxxxxxxxxxxxxxxx> <4D70A59A.2060603@xxxxxxxxxxxxxxxxxx>
Sender: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
On Fri, 2011-03-04 at 08:40 +0000, John Weekes wrote:
> On 3/3/2011 9:09 AM, Ian Campbell wrote:
> > On Thu, 2011-03-03 at 08:30 +0000, Keir Fraser wrote:
> >> >  Without the round-robin servicing, unfairness to the point of 
> >> > starvation is
> >> >  a distinct possibility.
> > Indeed. The following is a forward port of these patches from
> > 2.6.18-xen.hg:
> Thanks for working on this, Ian.
> I think that this has affected me several times, as I've noticed these 
> symptoms before (heavy disk access by a few domains, followed by several 
> Windows domUs mysteriously rebooting) -- as recently as Wednesday, in 
> fact. The I/O from rebooting domUs, unfortunately, makes everything even 
> worse, often leading to further Windows domUs crashing.
> I look forward to this being merged into stable-2.6.32. In the meantime, 
> I'll monkey around with git as you described and try to get a local 
> version working. I'm itching to test it out, since this also might 
> explain some of the occasional the longish disk delays that I see on 
> machines with a medium I/O load (in which the dom0 still responds 
> normally, but one of the domUs frequently stalls waiting for I/O).

This series was for more recent kernels. It seems to backport with
minimum fuss (switch from __this_cpu_(read|write) to __get_cpu_var).

I only compile tested this one file and haven't even booted it but the
result is available on top of xen.git xen/next-2.6.32 at:

The following changes since commit 892d2f052e979cf1916647c752b94cf62ec1c6dc:
  Jeremy Fitzhardinge (1):
        Merge commit 'v2.6.32.28' into xen/next-2.6.32

are available in the git repository at:

  git://xenbits.xen.org/people/ianc/linux-2.6.git irq-fairness-2.6.32

Ian Campbell (1):
      xen: events: Make last processed event channel a per-cpu variable.

Keir Fraser (3):
      xen: events: Clean up round-robin evtchn scan.
      xen: events: Make round-robin scan fairer by snapshotting each l2 word
      xen: events: Remove redundant clear of l2i at end of round-robin loop

Scott Rixner (1):
      xen: events: Process event channels notifications in round-robin order.

 drivers/xen/events.c |   80 +++++++++++++++++++++++++++++++++++++++++++++-----
 1 files changed, 72 insertions(+), 8 deletions(-)


Xen-devel mailing list