Re: [PATCH v3 5/9] firmware: arm_scmi: Make MBOX transport a standalone driver

[Cristian Marussi <cristian.marussi@arm.com>]

On Tue, Aug 06, 2024 at 02:12:39PM +0100, Sudeep Holla wrote:
> On Tue, Jul 30, 2024 at 02:33:14PM +0100, Cristian Marussi wrote:
> > Make SCMI mailbox transport a standalne driver that can be optionally
> > loaded as a module.
> > 

Hi Sudeep,

thanks for havig a look.

> > Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
> > ---
> > v2 --> v3
> > - fixed spacing in Kconfig
> > - updated Copyright
> > - use new params in DEFINE_SCMI_TRANSPORT_DRIVER
> > ---
> >  drivers/firmware/arm_scmi/Kconfig             |  4 +-
> >  drivers/firmware/arm_scmi/Makefile            |  3 +-
> >  drivers/firmware/arm_scmi/common.h            |  3 --
> >  drivers/firmware/arm_scmi/driver.c            |  3 --
> >  .../{mailbox.c => scmi_transport_mailbox.c}   | 47 +++++++++++++------
> >  5 files changed, 38 insertions(+), 22 deletions(-)
> >  rename drivers/firmware/arm_scmi/{mailbox.c => scmi_transport_mailbox.c} (87%)
> 
> I am happy with the changes in the series, they all look good. I wonder if
> it makes sense to move transport drivers into a separate folder
> drivers/firmware/arm_scmi/transport/{mailbox,smc,optee,virtio}.c
> 
> In scmi_transport_*.c, I see scmi is redundant and transport can be eliminated
> by moving all under the folder with that name. Thoughts ?
> 

Yes it is a possigility, not sure if there is any (trivial) drawback to
solve while movinf, but consider that the final name for the .ko LKM has to
have the scmi_transport prefix somehow to be able to live in the tree and distinguish
those LKMs from (vendor) protocols ones...so I would have anyway to craft an
scmi_ naming in the Makefile

Thanks,
Cristian