Re: [PATCH v10 0/7] ARM: support for Trusted Foundations secure monitor

[Alexandre Courbot <gnurou-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>]

On Thu, Nov 14, 2013 at 10:20 AM, Alex Courbot <acourbot-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org> wrote:
> On 11/14/2013 02:57 AM, Olof Johansson wrote:
>>
>> On Tue, Nov 12, 2013 at 6:14 PM, Alex Courbot <acourbot-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org> wrote:
>>>
>>> On 11/13/2013 05:38 AM, Olof Johansson wrote:
>>>>
>>>>
>>>> On Tue, Nov 12, 2013 at 12:26 PM, Stephen Warren <swarren-3lzwWm7+Weoh9ZMKESR00Q@public.gmane.org>
>>>> wrote:
>>>>>
>>>>>
>>>>> On 11/07/2013 03:11 AM, Alexandre Courbot wrote:
>>>>>>
>>>>>>
>>>>>> Just a set of small fixes to address the concerns expressed on v9 with
>>>>>> the
>>>>>> non-prefixed version DT properties. I hope there won't be a need for
>>>>>> an
>>>>>> eleventh (!) version. :P
>>>>>
>>>>>
>>>>>
>>>>> BTW, this version looks fine to me. On IRC, Olof said it looked OK to
>>>>> him. I'm just waiting to hear back from Olof/Russell whether I should
>>>>> merge this through the Tegra tree, or whether the first 1-3 patches
>>>>> should go through Russell's tree.
>>>>
>>>>
>>>>
>>>> I pinged Russell, and he brought up the fact that there were earlier
>>>> requests to move it to drivers/firmware. It would make sense to try to
>>>> get that done before merging, especially if you anticipate someone
>>>> using TF on 64-bit platforms.
>>>
>>>
>>>
>>> IIRC when we discussed this point your last comment was as follows:
>>
>>
>> Touche. :) Thanks for the reminder.
>>
>>> On Tue, Oct 29, 2013 at 6:55 AM, Olof Johansson <olof-nZhT3qVonbNeoWH0uzbU5w@public.gmane.org> wrote:
>>>>
>>>> I think we can probably merge this under arch/arm now, and when we
>>>> figure out what needs to be common with ARM64 we can move it out to a
>>>> good location. It might be that mostly just a header file with ABI
>>>> conventions needs to be shared, not actual implementation, for
>>>> example.
>>>
>>>
>>> So I thought we agreed on that. If in the end we prefer to move the ARM
>>> firmware interface into drivers/firmware, I'm fine with that too (Tomasz
>>> also confirmed he would be ok with it) but I wonder if that would not be
>>> somehow premature.
>>>
>>> Another worry of mine is that this might delay this patchset some more.
>>> Support for TF is one of the last remaining step towards making NVIDIA
>>> branded Tegra retail devices (SHIELD and TegraNote at the moment) run
>>> upstream directly. I missed 3.13, I'd like to make sure I won't miss
>>> 3.14.
>>> Would it be acceptable if we move the ARM firmware interface to a common
>>> place after this patchset is merged?
>>
>>
>> Well, as I already said I'm ok with things going into arch/arm to
>> start with, as long as Russell is. Once we see 64-bit needs for the
>> same we'll move it out -- it's not like it's a whole lot of code to
>> start with. But Russell has veto on the topic. :-)
>
>
> Thanks Olof. Russell, are you ok with the patchset in its current form? I
> can start moving the firmware interface out of arch/arm if that's what you
> want (there is no user outside of ARM at the moment, but as Olof pointed out
> that's not too much code) but I'd really like to see this series secured for
> 3.14.

Never mind, I have submitted a patch that moves firmware_ops to
drivers/firmware, that will hopefully settle this issue. Then maybe we
can finally flush this series as well (I will need to resubmit a new
version though).

Alex.