Re: [GIT PULL] fsi changes for 5.5

[Joel Stanley <joel@jms.id.au>]

On Thu, 7 Nov 2019 at 14:17, Greg KH <gregkh@linuxfoundation.org> wrote:
>
> On Thu, Nov 07, 2019 at 12:09:50PM +0000, Joel Stanley wrote:
> > Hi Greg,
> >
> > Here's a set of changes I'd like merged for 5.5. They've been well
> > tested in the openbmc tree over the past month or so as we've done
> > hardware bring up using them. Aside from the three fixes I applied
> > today they have seen time in linux-next too.
> >
> > This is the first time I've sent you a pull request, so please let me
> > know if you'd prefer it done differently.
> >
> > The following changes since commit 755b0ef68f1802c786d0a53647145a5a7e46052a:
> >
> >   fsi: aspeed: Clean up defines and documentation (2019-11-07 22:24:18 +1030)
>
> The pull request looks good, but some of the individual patches, I have
> questions on.  Also, a diffstat would be good so that I know I got it
> right for the next time you send this.

>
> As they aren't here in the emails, let me try to figure out how to
> respond:
>         - You have new dt bindings, yet no review from the DT
>           maintainers.

That was a mistake. I had it on the list but Rob must have not seen it
yet, as he's reviewed other patches of mine in the mean time.

>         - you move things around in sysfs, yet no documentation updates
>           happen

The documentation was incorrect, and these changes make the layout
closer to what is described in Documentation/ABI. I will include a
patch to clarify.

>         - in 0005-fsi-Add-ast2600-master-driver.patch you have lots of
>           dev_dbg() lines left that shoudl be dropped as that's what
>           ftrace is for
>         - you don't have any reviewers for some of these patches, that's
>           not good to stick in a pull request.

To give you some background, the code has all been closely reviewed as
it underwent development on the public openbmc list. We discussed
sending all of the patches with history preserved, but decided there
was very little value in having that history in the mainline kernel,
with the exception of patch 7 I mention below.

Unfortunately my co-developers are on leave or snowed under, so they
didn't get a chance to respond with r-b tags to the series I sent out
last week:

https://lore.kernel.org/linux-arm-kernel/20191101112905.7282-2-joel@jms.id.au/

>         - 0007-fsi-aspeed-Fix-OPB0-byte-order-register-values.patch does
>           not have a Fixes: tag, nor a stable@vger cc:, why not?

This patch is a fix for an earlier patch in the series. I did not
squash it in as I wanted the commit message to be part of history.

>         - 0010-fsi-fsi_master_class-can-be-static.patch has no changelog
>           text at all, which is not ok.
>
> Can you fix all of this up, and send it as a set of normal patches so at
> least I can review the things that do not have any other reviewers on
> it?

I will do this today.

Cheers,

Joel