These days in transplantation UCOSIII to STM32 project, good at the beginning, new mission, scheduling, interrupt handling are both good, some of the problems were solved smoothly, and then long to use the serial port is not when a serial port receives the data back in a few seconds to hang out,

What simulation and hardware testing done, struggled for several days, then also doubt yourself implanted UCOSIII have problems, ready to rewrite again only to find that the inside of the serial port interrupt function much receive data back there to write a semicolon, after removing the magic can! ? Wrote a C knows, semicolon just let the system know the end of one line of code, much more than just a semicolon theory a blank statement, should not hurt the refined, but the reality is very cruel, online also didn't find the related content,

In the process of programming keil did not give the error or warning, so can only be blank UCOSIII does not support this statement?

CodePudding user response:

Can generate cross-compilation file to see assembly code, the difference between a likely is a compiler BUG