Reason why non-volatile storage partition sector size has to be < UINT16_T MAX?

[vChavezB]

See this condition

https://github.com/zephyrproject-rtos/zephyr/blob/488fd89e2bd64c60354378fdec33b7e65c713d38/subsys/settings/src/settings_nvs.c#L369C2-L369C2

I guess it has to do that the maximum value expected for CONFIG_SPI_NOR_FLASH_LAYOUT_PAGE_SIZE is 65536?

[Laczen]

This is because NVS is internally using a location descriptor that only supports blocks that are smaller then 65536. It is a NVS limitation that cannot be removed without heavily modifying nvs.

[vChavezB]

Could you point out where in the subsys this is implemented (the location descriptor you mentioned). It would be nice to know to understand more about the NVS.

Best regards and thanks for the info.

[Laczen]

https://github.com/zephyrproject-rtos/zephyr/blob/main/subsys/fs/nvs/nvs_priv.h#L14-L22.

All uint32_t *addr used in the routines are formatted according to the above lines. It is known that nvs is unusable on devices with larger sector size (e.g. some stm32 devices).