drivers: flash: spi: Move to using select in Kconfig for SPI bus

[pdgendt]

• Updated to the latest policy to select the SPI bus for flash controllers
• Updated SPI flash samples to remove default selected symbols

[galak]

Marking DNM to allow some board maintainer feedback based on @de-nordic's comment.

[pdgendt]

Marking DNM to allow some board maintainer feedback based on @de-nordic's comment.

@galak Anyone in particular we should ping?

[github-actions]

This pull request has been marked as stale because it has been open (more than) 60 days with no activity. Remove the stale label or add a comment saying that you would like to have the label removed otherwise this pull request will automatically be closed in 14 days. Note, that you can always re-open a closed pull request at any time.

[pdgendt]

@galak is the DNM label still applicable?

[galak]

@henrikbrixandersen, @dleach02 , @danieldegrasse can you guys take a look

[galak]

@pdgendt looks like this needs a rebase to resolve a merge conflict.

[tejlmand]

currently blocking.

Although a discussion seems to have happened in #48707 then I was not consulted on such change and I need to take a closer look at this discussion because there are risks / pitfalls related to select when comparing to depends on.

I need to take a closer look at first #48707.

[carlescufi]

Architecture WG:

• depends on <BUS> vs select <bus>
• The <BUS> is only enabled if the corresponding DT_HAS_<>_ENABLED Kconfig option is enabled, which it is only if there's a Devicetree node enabled
• This was done for all sensors in the tree, see this commit
• @henrikbrixandersen mentions that they use another compatible whenever they extend an upstream driver, which then decouples the OoT implementation from the original one
• @gmarull mentions how, before @galak's commit was merged, defconfig files were unmaintainable, and so going back to the old approach is not really maintainable
• @mbolivar-nordic mentions that there are workarounds for sensors with multiple buses:

  zephyr/drivers/sensor/bme280/Kconfig

  Line 11 in 3454a25

  select I2C if $(dt_compat_on_bus,$(DT_COMPAT_BOSCH_BME280),i2c)

• @galak mentions that the very specific case of sensors on a particular bus applies here. A bus driver interface is what is being discussed here
• This was discussed in August: https://docs.google.com/document/d/13YxmRsaZ7u0tarhQ31Ca_MeHcg9M9e_GyHJqSYfnSpQ/edit#
• @galak to translate this into doc: drivers: sensor: Convert Kconfig bus 'depends on' to 'select' #48707
• @jfischer-no suggests using imply instead of select
• @henrikbrixandersen suggests revisiting the FORCE_ALT pattern, and instead suggests that using alternative compatibles would be a better idea

Dismissing as per architecture group review

[tejlmand]

A bit concerned on selecting higher level SPI from drivers.

One concern is the stuck-symbol which may occur when re-configuring.
Second concern is the select always kicking in, meaning users cannot get rid of a select as well as select overruling any depends on (a behavior which often surprises users)
There are some general design discussions / concerns as well, but would go to far to list everything here.

I acknowledge the limitation of current Kconfig / dts implementation wrt. scalability as described by df81fef

To accommodate concerns stated above we should see if a corresponding hidden (promptless) symbol should be introduced in addition to, e.g menuconfig SPI.
But that is considered outside the scope of this PR.

Approving with the concerns given above.

[carlescufi]

@pdgendt I've rebased the branch

[tejlmand]

Ref: #51598 (review)