[libc-commits] [libc] [llvm] [libc][math][c23] Improve rsqrtf16 function for targets without fp32 FPUs. (PR #160639)

[Anton Shepelev via libc-commits]

================
@@ -24,6 +24,237 @@
 namespace LIBC_NAMESPACE_DECL {
 namespace math {
 
+namespace rsqrtf16_internal {
+
+using FPBits = fputil::FPBits<float16>;
----------------
amemov wrote:

Done. Do we have any examples in the repo on how the ISA path can be optimized for float16? So far I only saw some examples of optimizations for x86 sqrt `long double`

https://github.com/llvm/llvm-project/pull/160639