Re: [PATCH v2 2/2] reset: Reset controller driver for Intel LGM SoC

[Martin Blumenstingl]

Hi Philipp,

On Thu, Oct 3, 2019 at 4:19 PM Philipp Zabel <pza@xxxxxxxxxxxxxx> wrote:
[...]
> > because the register layout was greatly simplified for the newer SoCs
> > (for which there is reset-intel) compared to the older ones
> > (reset-lantiq).
> > Dilip's suggestion (in my own words) is that you take his new
> > reset-intel driver, then we will work on porting reset-lantiq over to
> > that so in the end we can drop the reset-lantiq driver.
>
> Just to be sure, you are suggesting to add support for the current
> lantiq,reset binding to the reset-intel driver at a later point? I
> see no reason not to do that, but I'm also not quite sure what the
> benefit will be over just keeping reset-lantiq as is?
according to Chuan and Dilip the current reset-lantiq implementation
is wrong [0].
my understanding is that the Lantiq and Intel LGM reset controllers
are identical except:
- the Lantiq variant uses a weird register layout (reset and status
registers not at consecutive offsets)
- the bits of the reset and status registers sometimes don't match on
the Lantiq variant
- the Intel variant has a dedicated registers area for the reset
controller registers, while the Lantiq variant mixes them with various
other functionality (for example: USB2 PHYs)

> > This approach means more work for me (as I am probably the one who
> > then has to do the work to port reset-lantiq over to reset-intel).
>
> More work than what alternative?
compared to "fixing" the existing reset-lantiq driver (reset callback)
and then (instead of adding a new driver) integrating Intel LGM
support into reset-lantiq

> > I'm happy to do that work if you think that it's worth following this
> > approach.  So I want your opinion on this before I spend any effort on
> > porting reset-lantiq over to reset-intel.
>
> Reset drivers are typically so simple, I'm not quite sure whether it is
> worth to integrate multiple drivers if it complicates matters too much.
> In this case though I expect it would just be adding support for a
> custom .of_xlate and lantiq specific register property parsing?
yes, that's how I understand the Lantiq and Intel reset controllers:
- reset/status/assert/deassert callbacks would be shared across all variants
- register parsing and of_xlate are SoC specific


Martin


[0] https://www.spinics.net/lists/devicetree/msg305951.html