[PATCH 0/7] realtek: add HPE 1920 support

[Daniel Golle]

On Thu, Jul 28, 2022 at 03:27:14PM +0200, Sander Vanheule wrote:
> On Thu, 2022-07-28 at 15:10 +0200, Jan Hoffmann wrote:
> > > This adds support for three switches from the HPE 1920 series. It has
> > > been tested on HPE 1920-8G and HPE 1920-16G. Support for HPE 1920-24G
> > > is also included, as it uses the same board as the 16-port model.
> > > 
> > > The patch series depends on the firmware-utils patch adding the
> > > mkh3cimg and mkh3cvfs tools.
> > 
> > This patch series has just been merged. However, as mentioned in the 
> > cover letter, it depends on new tools in firmware-utils. As the 
> > corresponding patch series ("add tools for H3C devices") hasn't been 
> > merged yet, this won't actually build.
> 
> I've pushed a revert commit to avoid breaking the builds.
> 
> Can someone explain why this whole series needed to be merged so quickly? These *7* patches have
> only been posted on Saturday, and testing was performed for patches 2-4. At least patches 5-7
> could've remained unmerged awaiting some further comments.

As 5-7 depend on the actual device for testing they are unlikely to
receive great amounts of feedback, but are also unlikely to break
anything.

> 
> > 
> > Please tell me if I should have made this dependency clearer in any way, 
> > so I can avoid similar situations in the future.
> 
> It was plainly stated in the cover letter, I don't think you could have been any clearer about this
> dependency.

Yes, it was absolutely clear and I had it in my local firmware-utils
tree for which I used CONFIG_SRC_TREE_OVERRIDE.
The fact that the cover letter doesn't become part of git history was
part of the reason why I forgot about that before pushing.