2013-06-07 23:17:56

by luke.leighton

[permalink] [raw]
Subject: Re: [Arm-netbook] getting allwinner SoC support upstream (was Re: Uploading linux (3.9.4-1))

On Sat, Jun 8, 2013 at 12:09 AM, Dennis Lan (dlan)
<[email protected]> wrote:
>
>
> On Saturday, June 8, 2013, luke.leighton wrote:
>>
>> right - too many people contributed to this, input from jon smirl,
>> wookie, maxime, tomasz, henrik, i've made a start here and will
>> continue editing: this is notes for me to put forward an agenda for
>> discussion:
>>
>> http://hands.com/~lkcl/allwinner_linux_proposal.txt
>>
>> i'm setting a rule that each section *has* to have a list of clear
>> benefits, otherwise it'll have to be removed before it goes on to
>> their Directors.
>>
>> so - even if there are any allwinner engineers reading this who would
>> like something put forward please also speak up! :)
>>
>> l.
>>
>> _______________________________________________
>> linux-arm-kernel mailing list
>> [email protected]
>> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
>
>
>
> Hi luke
> I'm not a allwinner employer :-)$. but pretty much in the same position
> as they are.

thanks for chipping in.

> I'd like to add a few comments about the risk of adopting the device
> tree(from allwinner side)
> 1) current using fdt in bootloader(uboot) is not mature, I'm not saying
> about passing the fdt data to kernel.

mmmm.... fdt. could you provide some context here? what is it?
(apart from being a TLA)

> But the bootloader itself need information from device tree, say boot0/1
> phase (boot device type, DDR initialization...)
> since fdt is not ready, and SRAM space is very limited ... I'm afraid
> 'fex' may co-exist with device tree.
> still, they receive benefits if they can adopt device tree, at least
> minimal the kernel side migration effort
> Generally this info already been pointed out by steppen warren in previous
> email...

... which i have to admit i may have missed the significance of or
possibly just missed it entirely.

what's the main concern you have, here; what's the potential
solution, and what's the benefit of that potential solution, to
allwinner [direct or indirect]?

> 2) device tree may not been understood by third vendors (who previus produce
> shoes or ? :-$),

:)

> they are real old 'Fex' scheme user, they like edit the data in windows
> with dos format
> So, how to fill this gaps, make them happy? Creat another tool to handle
> device tree modification?
> Then it's another price they have to pay...

yehh... that kinda looks unavoidable, although it would ultimately
only inconvenience the developers of the proprietary firmware-flashing
tools [livesuite, phoenix] and so would be transparent to the
factories and so on. i've mentioned the idea of having an in-kernel
translation tool rather than an external (pre-runtime) one, i've yet
to get some feedback on that.

l.