[PATCH linux dev-6.5 v4 0/2] LTC4286 and LTC4287 driver support

[Patrick Williams]

Hello Joel,

On Tue, Dec 12, 2023 at 12:42:18PM +1030, Joel Stanley wrote:
> I have some
> new rules for getting the patches merged into openbmc:

Can you please update this wiki with what your current expectations are
then because these are not obviously reflected:

https://github.com/openbmc/linux/wiki/DevelopmentProcess

> 1. Do not send them for backporting to the dev-6.6 branch until they
> have been reviewed upstream. This means you have Reviewed-by or
> Acked-by tags on at least a majority of the patches in a series before
> you send them to the openbmc list.

> 2. Find reviewers for your upstream patches. Get other Facebook
> employees, get other openbmc contributors to review your patches. A
> good way to encourage others to review your patches is to first review
> thiers.

The submitters here are not Facebook employees (nit: there is also no such
thing as the company changed names 2 years ago now), but they are doing
work partially on behalf of Meta as their company is partnering with us to
develop new systems.

In this case, I don't think review has been an issue since Guenter
applied v7 to his tree.

> 3. When you do send the patches for backporting, include a
> justification in the cover letter for why they should be backported.
> For example: "These patches are merged upstream" or "the changes under
> active review, but we wish to have them in the openbmc tree because it
> has been ongoing for more than two weeks".

It would be nice for the wiki to give clear unbiased criteria here.
Currently it says:

   Patches for pre-upstream inclusion ... should be at least of a level
   approximately ready to submit upstream.

This implies, by my reading, that any commit sent upstream can in
parallel be sent to our 'dev-*' branches, since that is a stronger
criteria than you've listed.

Anyone can carry patches in their own downstream trees, so it isn't
obvious to me when it is deemed beneficial and not a burden to also
request a backport to the openbmc/linux branches.  I get the impression
that our backports requests are more on the burden end of the spectrum.

> Your latest upstream patchset produces a large number of warnings when
> I applied it to the 6.7-rc1 tree. This shows you are not compiling
> your changes before submitting.
> 
> Please engage with me and the other developers on this issue so we can
> help each other get the changes up to scratch and merged.
> 
> Cheers,
> 
> Joel

-- 
Patrick Williams
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.ozlabs.org/pipermail/openbmc/attachments/20231212/9121712c/attachment-0001.sig>