wolfBoot_update_trigger: Fix double-write-after-erase in NVM...WRITEONCE

[danielinux]

when wolfBoot_update_trigger is called on a device with NVM_ write-once workaround, it will cause an illegal second write to flash to set the UPDATING flag.

Instead, write the flag and the trailer in NVM_CACHE before committing to the flash.

Reported in zd19815

[Copilot]

Pull Request Overview

This PR fixes a double-write-after-erase issue for write-once NVM devices by staging the update flag and trailer in NVM_CACHE before performing a single flash write, and updates unit tests to reflect the new selected sector index.

• Updated unit tests to expect selection of sector 1 and verify erasure of bank 0.
• Refactored wolfBoot_update_trigger to erase the full sector and write flags/trailer from cache for write-once NVM.
• Removed redundant partition-state call and moved state updates into each branch.

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated no comments.

File	Description
tools/unit-tests/unit-nvm.c	Adjusted expected fresh sector from 0 to 1 and bank assertion.
tools/unit-tests/unit-enc-nvm.c	Updated expected fresh sector from 0 to 1 and bank assertion.
src/libwolfboot.c	Staged flags/trailer in cache for write-once NVM and updated erase size.

Comments suppressed due to low confidence (3)

src/libwolfboot.c:763

• [nitpick] Add a comment explaining the offset calculation and the adjustment under FLAGS_HOME to clarify why flags and trailer are positioned at this offset.

        uint32_t offset = SECTOR_FLAGS_SIZE;

tools/unit-tests/unit-nvm.c:269

• Introduce a dedicated unit test for the NVM_FLASH_WRITEONCE path to verify that flags and trailer are staged in cache and only one flash write occurs.

    /* Triggering update to set flags */