LLS: when turning off all LEDs on boot, skip LED-C Having LLS turn off LED-A and LED-B on boot is normally unnecessary (they should already be off in Iota), but it is harmless, hence this logic is kept for robustness. However, having LLS read-modify-write the BCICTL2 register (to turn off LED-C) creates a potential race condition with FCHG writes to this register, especially in the case when baseband switch-on is caused by VCHG and charging is expected to start right away. Furthermore, control of the charging LED itself (on those hw targets that have it) is the responsibility of the FCHG SWE, hence LLS should leave it alone.