[0/3] platform/x86: wmi: add WMI support to MODULE_DEVICE_TABLE()

Message

Mattias Jacobsson Jan. 19, 2019, 11:55 a.m. UTC

This patchset adds WMI support to MODULE_DEVICE_TABLE().

[PATCH 1/3]: prepare struct wmi_device_id
[PATCH 2/3]: add support
[PATCH 3/3]: update existing drivers to use MODULE_DEVICE_TABLE()

Mattias Jacobsson (3):
  platform/x86: wmi: move struct wmi_device_id to mod_devicetable.h
  platform/x86: wmi: add WMI support to MODULE_DEVICE_TABLE()
  platform/x86: wmi: use MODULE_DEVICE_TABLE() instead of MODULE_ALIAS()

 drivers/platform/x86/dell-smbios-wmi.c       |  2 +-
 drivers/platform/x86/dell-wmi-descriptor.c   |  2 +-
 drivers/platform/x86/dell-wmi.c              |  4 ++--
 drivers/platform/x86/huawei-wmi.c            |  3 +--
 drivers/platform/x86/intel-wmi-thunderbolt.c |  2 +-
 drivers/platform/x86/wmi-bmof.c              |  2 +-
 drivers/platform/x86/wmi.c                   |  2 +-
 include/linux/mod_devicetable.h              | 13 +++++++++++++
 include/linux/wmi.h                          |  5 +----
 scripts/mod/devicetable-offsets.c            |  3 +++
 scripts/mod/file2alias.c                     | 18 ++++++++++++++++++
 11 files changed, 43 insertions(+), 13 deletions(-)

Comments

Darren Hart Jan. 26, 2019, 9:06 p.m. UTC | #1

On Sat, Jan 19, 2019 at 12:55:52PM +0100, Mattias Jacobsson wrote:
> This patchset adds WMI support to MODULE_DEVICE_TABLE().

Hi Mattias,

Thanks for the patch series. I've reviewed and found no major issues - but what
I'm missing from this cover letter and each commit message is the why. What is
the problem this series is intended to address? This should be clear in the
commit messages so developers reading the git history have the necessary context
to understand why the change was made and the intent behind them were.

The only advantage that I see by the end of the series is removing the need for
driver authors to prefix the modules alias  with "wmi:" - which is an advantage
and avoids careless errors. Is that the only goal? It adds some complexity by
spreading the implementation across more files and more directories, so we need
to document the justification.

Thanks,

Mattias Jacobsson Jan. 27, 2019, 7:15 p.m. UTC | #2

Hi Darren,

On 2019-01-26, Darren Hart wrote:
> On Sat, Jan 19, 2019 at 12:55:52PM +0100, Mattias Jacobsson wrote:
> > This patchset adds WMI support to MODULE_DEVICE_TABLE().
> 
> Hi Mattias,
> 
> Thanks for the patch series. I've reviewed and found no major issues - but what
> I'm missing from this cover letter and each commit message is the why. What is
> the problem this series is intended to address? This should be clear in the
> commit messages so developers reading the git history have the necessary context
> to understand why the change was made and the intent behind them were.

Thanks for reviewing the patchset!

> 
> The only advantage that I see by the end of the series is removing the need for
> driver authors to prefix the modules alias  with "wmi:" - which is an advantage
> and avoids careless errors. Is that the only goal? It adds some complexity by
> spreading the implementation across more files and more directories, so we need
> to document the justification.

Now that you point it out I can definitely see that the reasoning isn't
explicitly stated. Sorry about that!

I've updated the commit messages for [PATCH 2/3] and [PATCH 3/3] in v2
of this patchset to clarify the "why".

> 
> Thanks,
> 
> -- 
> Darren Hart
> VMware Open Source Technology Center

Thanks,
Mattias