[U-Boot] [PATCH v2 0/45] Verified boot implementation based on FIT

[Simon Glass]

Hi Tom,

On Thu, Apr 25, 2013 at 11:51 AM, Simon Glass <sjg at chromium.org> wrote:
> Hi Tom,
>
> On Tue, Apr 23, 2013 at 7:50 AM, Tom Rini <trini at ti.com> wrote:
>> On Mon, Apr 22, 2013 at 06:44:53PM -0500, Kim Phillips wrote:
>>> On Sat, 20 Apr 2013 16:03:20 -0700
>>> Simon Glass <sjg at chromium.org> wrote:
>>>
>>> > On Mon, Apr 1, 2013 at 5:13 PM, Kim Phillips <kim.phillips at freescale.com> wrote:
>>> > > On Mon, 18 Mar 2013 16:51:20 -0700
>>> > > Simon Glass <sjg at chromium.org> wrote:
>>> > >
>>> > >> I have received a number of off-list comments - please do copy the list when
>>> > >> replying so that everyone can see your comments.
>>> > >
>>> > > I don't have time to fully review 45 patches, let alone the subject
>>> > > matter (e.g., no support for RSA in h/w, eh? ;), but I did notice
>>> > > some things that bugged me:
>>> > >
>>> > > Re: "[PATCH v2 04/45] libfdt: Add fdt_next_subnode() to permit easy subnode iteration":
>>> > >
>>> > > - Where's our libfdt maintainer?  libfdt patches should be submitted
>>> > > to the dtc project first.  It appears "[PATCH v2 40/45] libfdt: Add
>>> > > fdt_find_regions()" also suffers from this problem.
>>> >
>>> > The fdt_next_subnode() is a pretty trivial change. I will submit it to
>>> > the dtc project.
>>>
>>> Ideally we'd apply the patch directly from how it was applied to the
>>> upstream dtc project.
>>
>> And importantly, that's how we try and work this area in general.  Get
>> it upstream into dtc, mirror the change.  So lets get all of the fdt
>> related stuff there and pull it back down.
>
> Yes - the major patch was sent back around the same time as the
> verified boot series, along with an fdtgrep tool. I sent the
> fdt_next_subnode() patch earlier today.

OK this was accepted to dtc upstream in modified form. I will update
the series for U-Boot.

In case you want an update on the overall work, I have now split things into:

sandbox - http://patchwork.ozlabs.org/bundle/sjg/sandbox/ and I sent a
pull request if you want this
image1 - Allow images to work on sandbox, sent, needs this libfdt patch replaced
image2 - Introduce a common image_setup_linux() function, to send
image3 - image: Reduce code duplication and refactor, to send
vboot - Verified boot implementation based on FIT, final series, to send

I will do image1 in the next few days, and the rest hopefully next
week. I think it makes sense having vboot last.

Regards,
Simon