rpi_pico: Fix DTC warnings

[yonsch]

This PR fixes two DTC warnings that have been present for some time:

Warning (unique_unit_address_if_enabled): /soc/pin-controller@40014000: duplicate unit-address (also used in node /soc/gpio@40014000)

and:

unit address and first address in 'reg' (0x10000000) don't match for /soc/flash-controller@10000000/flash@0

The first was caused by the pinctrl and gpio bank sharing the same unit address. The RP2040 doesn't have an actual pin control unit, so it does not need a unit address.
The second was caused by a conflict of the unit address of the flash as defined by the SoC and the board. There wasn't a clear distinction between the flash controller and the flash itself. This was fixed for the SoC, board and driver.

[de-nordic]

Not an RPI pico expert and got some questions:

1. the DTS sets some chosen values that reference specific flash and flash controller, so should the code follow these chosen values?
2. or the chosen are set for Zephyr but the code can only support the specific flash and flash-controller so it is better to address them directly, rather than via chosen?

[yonsch]

@de-nordic The RP2040 uses an external flash, that is always mapped to 0x10000000. The HAL code addresses that directly. With Zephyr, I suppose using chosen to select that flash is more appropriate, so I fixed the driver as you suggested.

[de-nordic]

@de-nordic The RP2040 uses an external flash, that is always mapped to 0x10000000. The HAL code addresses that directly. With Zephyr, I suppose using chosen to select that flash is more appropriate, so I fixed the driver as you suggested.

I have accepted the solution as is. But still have some questions that may require some discussion and addressing.

So the HAL determines how this works, the DTS only describes the state for the Zephyr.
Does HAL use DTS to get these addresses or has them hardcoded somehow?
Will changing DTS will affect how HAL sees the flash, or it will only change how Zephyr (and Zephyr compilation) sees the underlying hardware?

If HAL exports the configuration via some header, shouldn't it be used in the driver rather than the DTS?
Assuming that HAL does not use DTS: this is RPI specific driver, for SoC integral controller, shouldn't the driver use the same identifiers as HAL does?
Just wondering...