Currently we compile the kernel with -std=gnu99. Someone recently
added some code that uses the u"foo" syntax for UTF-16 literals, which
is new in C11. With -std=gnu99, gcc accepts this syntax but clang
rejects it -- clang requires -std=c11 or -std=gnu11.
We could switch to L"foo" like is being used elsewhere, but it's not
exactly correct -- the data in question really are fixed to be UTF-16,
not some arbitrary interpretation of wide characters. We may also
want to use other C11 features such as _Atomic in the future.
Any objections to compiling the kernel with -std=gnu11 instead?
What are the reasons why we cannot use