From: Wei Liu Subject: Re: [PATCH RFC] xen/balloon: check whether a page is pointed to scratch page MFN Date: Thu, 5 Sep 2013 15:27:51 +0100 Message-ID: <20130905142751.GD14104@zion.uk.xensource.com> References: <1378375219-28193-1-git-send-email-wei.liu2@citrix.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from mail6.bemta14.messagelabs.com ([193.109.254.103]) by lists.xen.org with esmtp (Exim 4.72) (envelope-from ) id 1VHaYY-0005RM-9T for xen-devel@lists.xenproject.org; Thu, 05 Sep 2013 14:29:10 +0000 Content-Disposition: inline In-Reply-To: List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Stefano Stabellini Cc: xen-devel@lists.xenproject.org, Wei Liu , david.vrabel@citrix.com List-Id: xen-devel@lists.xenproject.org On Thu, Sep 05, 2013 at 03:21:32PM +0100, Stefano Stabellini wrote: > On Thu, 5 Sep 2013, Wei Liu wrote: > > Now that balloon pages might have two kind of P2M entries, a) > > INVALID_P2M_ENTRY when that page is never used, b) mfn of one of the > > many ballooned scratch pages. Refelct this in increase_reservation to > > avoid hitting BUG_ON. > > > > Signed-off-by: Wei Liu > > --- > > drivers/xen/balloon.c | 21 ++++++++++++++++++++- > > 1 file changed, 20 insertions(+), 1 deletion(-) > > > > diff --git a/drivers/xen/balloon.c b/drivers/xen/balloon.c > > index 3101cf6..63a7e5e 100644 > > --- a/drivers/xen/balloon.c > > +++ b/drivers/xen/balloon.c > > @@ -305,6 +305,24 @@ static enum bp_state reserve_additional_memory(long credit) > > } > > #endif /* CONFIG_XEN_BALLOON_MEMORY_HOTPLUG */ > > > > +static bool p2m_is_ballooned_scratch_page(unsigned long pfn) > > +{ > > + int cpu; > > + unsigned long bpfn; > > + struct page *bpage; > > + > > + for_each_possible_cpu(cpu) { > > + bpage = per_cpu(balloon_scratch_page, cpu); > > + if (!bpage) > > + continue; > > + bpfn = page_to_pfn(bpage); > > + if (pfn_to_mfn(pfn) == pfn_to_mfn(bpfn)) > > Even though it is true that this function doesn't need to be very fast, > we might as well try to write in an efficient way. > For example it might be worth calculating pfn_to_mfn(pfn) just once > outside the loop? > Good point.