[libc][math][c23] Clang 11 fails to build some _Float16 math functions for AArch64

[overmighty]

See:

• https://lab.llvm.org/buildbot/#/builders/138/builds/67159
• https://lab.llvm.org/buildbot/#/builders/223/builds/43033
• [libc][math] Temporarily disable nexttowardf16 on aarch64 due to clang-11 bug. #94569
• https://lab.llvm.org/buildbot/#/builders/138/builds/67337
• https://lab.llvm.org/buildbot/#/builders/223/builds/43215
• [libc][math][c23] Temporarily disable fmodf16 on AArch64 #94813
• https://lab.llvm.org/buildbot/#/builders/138/builds/67428
• https://lab.llvm.org/buildbot/#/builders/223/builds/43306
• [libc][math][c23] Temporarily disable modff16 on AArch64 #94972

cc @lntue

[nickdesaulniers]

We should probably upgrade our minimally supported version of clang at this point.