Re: [PATCH v2 06/24] clk: zte: Add ZX279128S CRM clock and reset driver

[Navid Ghahremani <ghahramani.navid@gmail.com>]

Hi Philipp,

Thanks for the review. The next version splits this up like Stefan
asked. The resets move out of drivers/clk into a reset driver in
drivers/reset that uses the CRM regmap and is probed from an MFD
parent, like Stefan's zx297520v3 reset driver that you reviewed. Your
points apply to that driver.

Philipp Zabel wrote:
> Don't select RESET_SIMPLE without using it.

The new driver won't select it.

> Why share the lock? Each reset controller has its own register.

The new driver doesn't share a lock, the regmap has its own.

> This check is unnecessary, fwnode_reset_simple_xlate() already checked
> this before the reset control was created.

That check is gone in the new driver.

> Why is this initialized from CLK_OF_DECLARE_DRIVER()?
> Surely USB resets are not required early?

Only because the reset code was part of the early clock driver. Nothing
needs these resets early, and the new driver is a normal platform
driver.

> This is exactly the same as zx_reset_data, why duplicate it?

The two controllers will share one data structure and one set of ops.

Best regards,
Navid