[clang] [llvm] [NVPTX] Add conversion intrinsics from/to fp8 types (e4m3, e5m2) (PR #102969)

[Artem Belevich via cfe-commits]

================
@@ -968,6 +971,39 @@ __device__ void nvvm_cvt_sm80() {
   // CHECK: ret void
 }
 
+// CHECK-LABEL: nvvm_cvt_sm89
+__device__ void nvvm_cvt_sm89() {
+#if __CUDA_ARCH__ >= 890
+  // CHECK_PTX81_SM89: call i16 @llvm.nvvm.ff.to.e4m3x2.rn(float 1.000000e+00, float 1.000000e+00)
+  __nvvm_ff_to_e4m3x2_rn(1, 1);
----------------
Artem-B wrote:

Nit: 1->1.0f. While implicit conversion works, it looks a bit odd when the check line matches `float 1.000000e+00`

https://github.com/llvm/llvm-project/pull/102969