RE: [PATCH 1/4] drivers/bus: Added Freescale Management Complex APIs

[Stuart Yoder]

> -----Original Message-----
> From: Wood Scott-B07421
> Sent: Thursday, September 18, 2014 6:29 PM
> To: Yoder Stuart-B08248
> Cc: Phillips Kim-R1AAHA; Alexander Graf; Rivera Jose-B46482; gregkh@xxxxxxxxxxxxxxxxxxx; arnd@xxxxxxxx;
> linux-kernel@xxxxxxxxxxxxxxx
> Subject: Re: [PATCH 1/4] drivers/bus: Added Freescale Management Complex APIs
> 
> On Thu, 2014-09-18 at 18:13 -0500, Yoder Stuart-B08248 wrote:
> >
> > > -----Original Message-----
> > > From: Kim Phillips [mailto:kim.phillips@xxxxxxxxxxxxx]
> > > Sent: Thursday, September 18, 2014 3:23 PM
> > > To: Alexander Graf
> > > Cc: Rivera Jose-B46482; <gregkh@xxxxxxxxxxxxxxxxxxx>; <arnd@xxxxxxxx>; <linux-kernel@xxxxxxxxxxxxxxx>;
> Yoder
> > > Stuart-B08248; Wood Scott-B07421; <linuxppc-release@xxxxxxxxxxxxxxxxxxx>
> > > Subject: Re: [PATCH 1/4] drivers/bus: Added Freescale Management Complex APIs
> > >
> > > On Thu, 18 Sep 2014 15:14:03 +0200
> > > Alexander Graf <agraf@xxxxxxx> wrote:
> > >
> > > > > Am 18.09.2014 um 06:17 schrieb German Rivera <German.Rivera@xxxxxxxxxxxxx>:
> > > > >
> > > > >> On 09/15/2014 06:44 PM, Kim Phillips wrote:
> > > > >> On Thu, 11 Sep 2014 12:34:21 -0500
> > > > >> "J. German Rivera" <German.Rivera@xxxxxxxxxxxxx> wrote:
> > > > >>
> > > > >>> From: "J. German Rivera" <German.Rivera@xxxxxxxxxxxxx>
> > > > >>>
> > > > >>> APIs to access the Management Complex (MC) hardware
> > > > >>> module of Freescale LS2 SoCs. This patch includes
> > > > >>> APIs to check the MC firmware version and to manipulate
> > > > >>> DPRC objects in the MC.
> > > > >>>
> > > > >>> Signed-off-by: J. German Rivera <German.Rivera@xxxxxxxxxxxxx>
> > > > >>> Signed-off-by: Stuart Yoder <stuart.yoder@xxxxxxxxxxxxx>
> > > > >>> ---
> > > > >>>  drivers/bus/fsl-mc/dpmng.c         |   93 +++++
> > > > >>>  drivers/bus/fsl-mc/dprc.c          |  504 +++++++++++++++++++++++
> > > > >>>  drivers/bus/fsl-mc/fsl_dpmng_cmd.h |   83 ++++
> > > > >>>  drivers/bus/fsl-mc/fsl_dprc_cmd.h  |  545 +++++++++++++++++++++++++
> > > > >>>  drivers/bus/fsl-mc/fsl_mc_sys.c    |  237 +++++++++++
> > > > >>>  include/linux/fsl_dpmng.h          |  120 ++++++
> > > > >>>  include/linux/fsl_dprc.h           |  790 ++++++++++++++++++++++++++++++++++++
> > > > >>>  include/linux/fsl_mc_cmd.h         |  182 +++++++++
> > > > >>>  include/linux/fsl_mc_sys.h         |   81 ++++
> > > > >>>  9 files changed, 2635 insertions(+)
> > > > >>>  create mode 100644 drivers/bus/fsl-mc/dpmng.c
> > > > >>>  create mode 100644 drivers/bus/fsl-mc/dprc.c
> > > > >>>  create mode 100644 drivers/bus/fsl-mc/fsl_dpmng_cmd.h
> > > > >>>  create mode 100644 drivers/bus/fsl-mc/fsl_dprc_cmd.h
> > > > >>>  create mode 100644 drivers/bus/fsl-mc/fsl_mc_sys.c
> > > > >>>  create mode 100644 include/linux/fsl_dpmng.h
> > > > >>>  create mode 100644 include/linux/fsl_dprc.h
> > > > >>>  create mode 100644 include/linux/fsl_mc_cmd.h
> > > > >>>  create mode 100644 include/linux/fsl_mc_sys.h
> > > > >>
> > > > >> the fsl prefix in the filename fsl_dpmng_cmd.h is redundant with
> > > > >> its directory name fsl-mc/.  Note that I find dashes ('-') in
> > > > >> filenames make them easier to type: is there a reason we're using
> > > > >> underscores here?
> > > > > This is a convention that we decided early on '-' for directory names
> > > > > and '_' for file names.
> > >
> > > based on what?
> >
> > We looked at how generally files were named in kernel source.
> >
> > For what it's worth in my 3.16 branch:
> >
> > $ find drivers/ -type f | grep '-' | wc -l
> > 4308
> > $ find drivers/ -type f | grep '_' | wc -l
> > 6507
> >
> > ...it seems that there are far more files named with underscores.  If Greg
> > or Arnd wants the files renamed, fine, but other than that I see no reason
> > to change this.
> 
> 60% isn't "far more".  Regardless of whether we change this (dashes are
> easier to read, easier to type, and are what English normally uses --
> underscores are for when you can't use a dash for technical reasons),
> let's not start making rules based on weak statistical inferences.
> 
> Plus, why limit the search to drivers?

It's a driver and we looked at the precedent/conventions used by
other drivers.

> E.g. if you looked at arch/
> you'd reach the opposite conclusion.

I guess if you look at the kernel as a whole it's almost exactly
50/50.  So, don't understand how anyone can say it is _clear_
that underscores are better.

Stuart