Posted on October 30, 2016 at 18:47Hi,
The latest STM32L0Cube UART Hal pulls in unnecessary floating point code when initializing the UART. The culprit is the __DIV_LPUART macro, which contains a floating point number (256.0). When replaced with the version from STM32L4Cube, it no longer pulls in floating point libraries.
The extra code takes up at least 6KB of extra code space, which is a lot for the smaller L0's.
Tony
