--- Quote Start ---
Assume there is a Carry chain, the first cin and the last cout must be floating according your comments? If it is true, how can i use the chain? becasue i must input my signal into this chain!!! Thanks for you reply, anyway!
--- Quote End ---
Obviously, floating cout isn't a problem at all and will be usually observed at the end of a carry chain.
Floating cin doesn't happen in real FPGA hardware. Consider that carry and carry_sum are abstract primitives that don't reflect exactly hardware properties. This causes some problems when describing actual hardware. The real hardware structure can be best seen in device manual and in the resource property editor.
In real FPGA hardware, at the start of the carry chain there's an LE that doesn't use cin for the cout expression. If I remember right, it can be described in carry primitive terms with a constant input to cin. But as said, I didn't use the carry primitives yet.
Honored Contributor
