[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20171013175927.GE17578@cbox>
Date: Fri, 13 Oct 2017 19:59:27 +0200
From: Christoffer Dall <cdall@...aro.org>
To: Auger Eric <eric.auger@...hat.com>
Cc: eric.auger.pro@...il.com, linux-kernel@...r.kernel.org,
kvm@...r.kernel.org, marc.zyngier@....com,
peter.maydell@...aro.org, andre.przywara@....com,
wanghaibin.wang@...wei.com, wu.wubin@...wei.com
Subject: Re: [PATCH v2 08/10] KVM: arm/arm64: vgic-its: free caches when
GITS_BASER Valid bit is cleared
On Fri, Oct 13, 2017 at 05:34:44PM +0200, Auger Eric wrote:
> Hi Christoffer,
>
> On 13/10/2017 17:19, Christoffer Dall wrote:
> > On Wed, Sep 27, 2017 at 03:28:38PM +0200, Eric Auger wrote:
> >> When the GITS_BASER<n>.Valid gets cleared, the data structures in
> >> guest RAM are not provisionned anymore. The device, collection
> >
> > provisioned
> >
> > (but did you really mean valid?)
> well the GITS_BASER<n>.Valid bit is reset, meaning the tables may be
> freed / un-provisioned by the driver. I can use valid though.
To me, the provision is the act of establishing something, but once it's
provisioned it just is, and then it isn't anymore. So your sentence
"...not provisioned anymore" tells me that there was a process where
someone was constantly establishing new data sructures in guest ram, and
the process stops. But what I think you want to say is that there were
data structures in place, but when the bit gets cleared, those data
structures are no longer used, i.e. they are no longer valid.
This is really just a wording thing, and I'm not a native English
speaker, but looking up provisioning in the dictionary seems to support
my claim here.
Thanks,
-Christoffer
Powered by blists - more mailing lists