Thread: "integer operation result is out of range" warning? Can someone explain why?

I'm working on an embedded project for an ARM based microcontroller. In one function, I'm casting six 1-byte integers into two words with 0xFFFF to make up the empty space. The lines are below:

Code:

report_addr[0] = (0xFFFF)<<16 | p_adv_report->peer_addr.addr[0]<<8 | p_adv_report->peer_addr.addr[1];

report_addr[1] = p_adv_report->peer_addr.addr[2]<<24 | p_adv_report->peer_addr.addr[3]<<16 | p_adv_report->peer_addr.addr[4]<<8 | p_adv_report->peer_addr.addr[5];

report_addr is an array of uint32.
peer_addr.addr is an array of uint8.

My compilers is giving me this warning:
#61-D: integer operation result is out of range.

Seems like there's exactly two bytes of data fitting into a 2-byte word. Why is this spitting out a warning? I'd like the project to build without the warning, if possible.

What 'type' is 0xffff?? Therein lies he issue. Surprised you didn't work that out yourself.

Originally Posted by Hobbit

What 'type' is 0xffff?? Therein lies he issue. Surprised you didn't work that out yourself.

I declared a uint16 variable with the hex value, then used that in the casting operation. Now, no warning.

Still learning... Thought I could simply provide the hex value rather than have to make a variable for it.

You simply needed to change 0xffff to 0xffffU.

The U tells the compiler that 0xffff is unsigned.

That worked, too. Thank you.