The divide_by_zero is meant to run if both __aarch64__ and __mips__ are not defined, meaning that the test is being compiled on any other architecture. As it is now, the test is being compiled also on __aarch64__ architecture, as !defined(__mips__) is true.
The divide_by_zero is meant to run if both
__aarch64__and__mips__are not defined, meaning that the test is being compiled on any other architecture. As it is now, the test is being compiled also on__aarch64__architecture, as!defined(__mips__)is true.