On Fri, May 19, 2023 at 02:45:26PM +0200, Michal Simek wrote:
Hi Ilias,
On 5/19/23 14:21, Ilias Apalodimas wrote:
HI Jassi, Michal
Based on the discussion we had on the dt bindings, I am personally ok with the notion of having those defined internally until we can prove it makes sense for the to be sent to the dt-schema.
In the future we need to strip those from U-Boot, before we hand over the DR to the OS, but this is a problem that already exists regardless of this patchset.
Jassi, Michal had some review comments. Are you going to send a v5 with those fixed?
We created this patch for dt-schema and if there is no comment I will be sending it to Rob to get merge.
lore.kernel.org/r/ca0715934133dbce6a5a3fd91483e0af92ea8ac6.1683815597.git.michal.simek@amd.com
As I said I am fine with having it in u-boot tree only but that node has to be removed to pass certification. It means I would expect that we will solve it together not later on because it will catch us very quickly. If you have code in EFI to do that it, let's just do it together to be able to pass IR2.0 requirements without waiting for another patch.
So I think in sum, I think the path today is to see if Rob will accept that patch and if not, either rework based on feedback or if the feedback is "this is U-Boot centric atm, not appropriate" we'll go with the path of removing nodes before the OS gets the tree, which I think is a generic enough bit of framework that we'll need regardless, for 2.0.