mirror of
https://github.com/intel/llvm.git
synced 2026-01-26 20:23:39 +08:00
0b72899f6d
This patch refactors the logic to define each component of the `math_errhandling` macro. It assumes that math error handling is supported by the target and the C library unless otherwise disabled in the preprocessor logic. In addition to the refactoring, the support for error handling via exceptions is explicitly disabled for Arm targets with no FPU, that is, where `__ARM_FP` is not defined. This is because LLVM libc does not provide a floating-point environment for Arm no-FP configurations (or at least one with support for FP exceptions).