Eliminate unnecessary loop in _cap_check()

[freqlabs]

Summary

Calling cap_rights_contains() several times with the same inputs is not going to produce a different output.

For context, here is the definition of cap_rights_contains() from sys/kern/subr_capability.c:

bool
cap_rights_contains(const cap_rights_t *big, const cap_rights_t *little)
{
        unsigned int i, n;

        assert(CAPVER(big) == CAP_RIGHTS_VERSION_00);
        assert(CAPVER(little) == CAP_RIGHTS_VERSION_00);
        assert(CAPVER(big) == CAPVER(little));

        n = CAPARSIZE(big);
        assert(n >= CAPARSIZE_MIN && n <= CAPARSIZE_MAX);

        for (i = 0; i < n; i++) {
                if ((big->cr_rights[i] & little->cr_rights[i]) !=
                    little->cr_rights[i]) {
                        return (false);
                }
        }

        return (true);
}

Test Plan

Test early. Test often.

[ed]

Interesting. I suspect an older version of this code hand-rolled cap_rights_contains() and was incorrectly refactored?

[freqlabs]

In D7369#153325, @ed wrote:

Interesting. I suspect an older version of this code hand-rolled cap_rights_contains() and was incorrectly refactored?

It sure looks that way, doesn't it. I dug through the logs to try to pinpoint where the refactor happened, but r255219 introduces the loop just like that. Lost artifact of development, I guess.

[freqlabs]

The variable i is not necessary, either.