LadOS | Install and setup scripts for LadOS | Configuration Management library

I'm not entirely sure why your original didn't work assuming int is the same size as int32_t for your compiler. If your vertice1->nombre is less than vertice2->nombre then the result of vertice1->nombre-vertice2->nombre will be a large unsigned value that is out of range for a 32-bit int. Most compilers will just map the out-of-range values to negative numbers although the actual result is implementation-defined.

When subtracting unsigned int values, a "negative" difference will end up as a large unsigned int value that is out-of-range for an int, so the result will be implementation-defined. For a qsort or bsearch comparison function, only the sign (positive, negative, or zero) of the returned value is used, so an implementation-defined result can be avoided by always returning -1 for a "negative" difference, 1 for a "positive" difference, or 0 for no difference. There are (at least) three ways to accomplish that :–

1. using if statements: