From mboxrd@z Thu Jan 1 00:00:00 1970 From: Julien Grall Subject: Re: [PATCH v4 33/33] docs/misc: arm: Add documentation about non-PCI passthrough Date: Thu, 09 Apr 2015 14:36:41 +0100 Message-ID: <55268069.9080802@gmail.com> References: <1426793399-6283-1-git-send-email-julien.grall@linaro.org> <1426793399-6283-34-git-send-email-julien.grall@linaro.org> <1427802939.2115.109.camel@citrix.com> <551A9B2C.5030204@linaro.org> <1427808109.2115.129.camel@citrix.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from mail6.bemta5.messagelabs.com ([195.245.231.135]) by lists.xen.org with esmtp (Exim 4.72) (envelope-from ) id 1YgCdr-0005Ow-4D for xen-devel@lists.xenproject.org; Thu, 09 Apr 2015 13:37:11 +0000 Received: by wiun10 with SMTP id n10so98384630wiu.1 for ; Thu, 09 Apr 2015 06:37:09 -0700 (PDT) In-Reply-To: <1427808109.2115.129.camel@citrix.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Ian Campbell , Julien Grall Cc: xen-devel@lists.xenproject.org, tim@xen.org, stefano.stabellini@citrix.com List-Id: xen-devel@lists.xenproject.org On 31/03/15 14:21, Ian Campbell wrote: > On Tue, 2015-03-31 at 14:03 +0100, Julien Grall wrote: > >>>> + passthrough { >>>> + compatible = "simple-bus"; >>>> + ranges; >>>> + #address-cells = <2>; >>>> + #size-cells = <2>; >>>> + mac0: ethernet@10000000 { >>>> + compatible = "calxeda,hb-xgmac"; >>>> + reg = <0 0x10000000 0 0x1000>; >>>> + interrupts = <0 80 4 0 81 4 0 82 4>; >>> >>> Tabs vs spaces? >> >> I think so. I will fix it. >> >>> >>> Is it worth making reference to looking at the host entry and copying >>> some of the relevant bits of that, or would that be more confusing? >> >> I don't understand the question here. > > Should we give people a hint that to discover the contents of this node > they should refer to the equivalent entry in the host db. I don't think it's necessary. In the current state, DT passthrough is very primitive and only a user who is aware of device tree concept can use it (some properties can't be added in the DT...). >> >>>> + }; >>>> + }; >>>> +}; >>>> + >>>> +Note: >>>> + * The interrupt-parent property will be added by the toolstack in the >>>> + root node; >>>> + * The properties compatible, ranges, #address-cells and #size-cells >>>> + in /passthrough are mandatory. >>> >>> The following properties are mandatory within the /passthrough node: >>> * compatible >>> * ranges >>> * ... >>> >>> Are the values of any of them mandatory, i.e. simply-bus or empty >>> ranges? >> >> "simple-bus" is mandatory. The empty ranges not. > > It's worth saying the former I think. > > For the latter a non-empty ranges would require exciting calculations > and changes to the iomem parameter to match etc, so perhaps recommend > that it be empty? I think the user is smart enough to decide which solution is better for him. Regards, -- Julien Grall