[libcxx-commits] [libcxx] [libcxx][test][z/OS] Fix hermite.pass.cpp for HEX float (PR #101019)

[via libcxx-commits]

================
@@ -26,7 +26,12 @@
 
 #include "type_algorithms.h"
 
-inline constexpr unsigned g_max_n = 128;
+inline constexpr unsigned g_max_n =
+#if !(defined(__MVS__) && !defined(__BFP__))
+    128;
+#else
+    39;
+#endif
----------------
PaulXiCao wrote:

Can we find a way to implement this more reliable to also work for other special architectures?

E.g. making use of `std::numeric_limits<Real>::max_exponent10`.
https://en.cppreference.com/w/cpp/types/numeric_limits/max_exponent10

Maybe via a function similar to 
```cpp
template <class Real>
constexp unsigned get_maximal_order() {
  if constexpr (..max_exponent10 < ...)
    return ...;
  else ...
}
```

https://github.com/llvm/llvm-project/pull/101019