One workaround is to change the negative constant in the assembly file to its unsigned equivalent and recompile the assembly file. In this case, the option to keep the assembly file must be used. The negative constant causing the compiler error should be converted to its binary representation, and that representation converted to an unsigned number.

A load of a 64-bit source constant such as the following example, which can be broken into 2 16-bit constants for bits 0-15 and 32-47 (bits 16-31 and 48-64 are 0), may incorrectly emit a negative constant in the assembly file, causing the assembler to issue an error. For example: unsigned long long x = 0xca0600006274;