pointers

The rationale for the conversion rules for pointers to pointers is based on a similar sort of problem.

With an intrusive linked list you pay the memory cost of having node pointers all the time.

Consequently pointers can be given out, and the kernel knowns if the app has tampered with them.

The pointers in the return value are valid as long as their particular descendant of the function parameter isn't deleted.

Strings are a good example of this: short strings can be tagged pointers, and long strings can be dynamically allocated.

Threads do share pointers and that's why they are more suitable to in memory data analysis/manipulation.

There are interesting pointers[2] as well as a part of a PhD thesis[3] that covers it.

It only seems messed up because you know about pointers, which are an implementation detail.

If the caller was responsible, the return value would be a vector of smart pointers.

The housekeeping struct has pointers for both the original function and the replaced function.

The linked articles in this LWN article (and the article itself) give some pointers.

But then they want to hire somebody and not even have to give him a few pointers, because that's too vulgar for them?

The nice thing is that for pointers, tagging is "free", for the lowest 4 bits of a 16-byte aligned pointer are not used anyway.

To consider function types, or unjustified assumptions about bitfield unions, or use of parentheses to control nesting of arrays and pointers in declaration "strange" one must be averse to the C language to the point of intolerance.