Re: [GIT PULL] platform-drivers-x86 for 4.15-1

From: Darren Hart
Date: Mon Nov 20 2017 - 14:06:11 EST


On Sat, Nov 18, 2017 at 10:37:55AM -0800, Linus Torvalds wrote:
> On Sat, Nov 18, 2017 at 10:09 AM, Andy Shevchenko
> <andriy.shevchenko@xxxxxxxxxxxxxxx> wrote:
> >
> > Here is the collected material against Platform Drivers x86 subsystem.
> > It's rather bit busy cycle for PDx86, mostly due to Dell SMBIOS driver
> > activity.
>
> So I note that you seem to use the same summary script that Darren used.
>
> Guys, the "merge summary" should be human-readable, not just a
> slightly munged version of "git shortlog" that has been sorted by
> first word.
>
> Feel free to use that munged version as a base-line, but don't just
> send me automated merge messages.
>
> The whole point of a summary is to summarize. And merge messages
> should be legible, not "this is all the details that you could have
> just used 'git shortlog' to get".

Hi Linus,

Back in the 4.2 timeframe, platform-drivers-x86-v4.2-2 specifically, I
started adding my pull request commentary to the tag directly and the
pull requests themselves tended to have little or no information beyond
that. I didn't see a clear division between what should go in the pull
request email body and what should be in the tag, this kept all the
information about the pull request together in the tag.

Andy's pull request follows this pattern, with the text of the tag
opening with the summary and context relevant to the pull request before
the munged shortlog:

--- 8< ---
For this cycle we have quite an update for the Dell SMBIOS driver
including WMI work to provide an interface for SMBIOS tokens via sysfs
and WMI support for 2017+ Dell laptop models. SMM dispatcher code is
split into a separate driver followed by a new WMI dispatcher.
The latter provides a character device interface to user space.

The pull request contains a merge of immutable branch from Wolfram Sang
in order to apply a dependent fix to the Intel CherryTrail Battery
Management driver.

Other Intel drivers got a lot of cleanups. The Turbo Boost Max 3.0
support is added for Intel Skylake.

Peaq WMI hotkeys driver gets its own maintainer and white list of
supported models.

Silead DMI is expanded to support few additional platforms.

Tablet mode via GMMS ACPI method is added to support some ThinkPad
tablets.

Two commits appear here which were previously merged during the
v4.14-rcX cycle:

- d7ca5ebf2493 platform/x86: intel_pmc_ipc: Use devm_* calls in driver probe function
- e3075fd6f80c platform/x86: intel_pmc_ipc: Use spin_lock to protect GCR updates
--- 8< ---

Do you not want to see this in the tag and prefer we move it to the GIT
PULL email body? If so, what do you want to see in the tag message?

Or, is the above not what you're looking for at all? And if not, what
are you looking for instead?

>
> When you can't even be bothered to fix up obvious issues like
>
> (a) "yeah, I just sorted things alphabetically without any human interaction":
>
> intel-wmi-thunderbolt:
> - Silence error cases
>
> MAINTAINERS:
> - Add entry for the PEAQ WMI hotkeys driver
>
> mlx-platform:
> - make a couple of structures static
>
> (b) "yeah, my summary script splits the overview at a colon, so when
> there is no colon it doesn't work":
>
> Add driver to force WMI Thunderbolt controller power status:
> - Add driver to force WMI Thunderbolt controller power status
>
> it's good that you have a fairly unified commit message model and can
> do these summaries, but it's bad when you then don't do much of any
> _human_ summary at all or even check the result for sanity.

Agreed. Andy, we need to read through everything we put into that pull
request and adjust to make sense. Pull Requests are not automated, the
scripts are there to help us avoid human error, and be as consistent as
possible.

--
Darren Hart
VMware Open Source Technology Center