[clang] [llvm] [HLSL] implement elementwise firstbithigh hlsl builtin (PR #111082)

[Steven Perron via cfe-commits]

================
@@ -0,0 +1,37 @@
+; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv-unknown-unknown %s -o - | FileCheck %s
+; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv-unknown-unknown %s -o - -filetype=obj | spirv-val %}
----------------
s-perron wrote:

If you look at FindSMsb and FindUMsb in https://registry.khronos.org/SPIR-V/specs/unified1/GLSL.std.450.html, it says:

> This instruction is currently limited to 32-bit width components.

For DXC, we opted to just not implement the 64-bit version. We can choose to do something different in Clang. See https://github.com/microsoft/DirectXShaderCompiler/issues/4702 for that discussion.

This is an issue for 16-bit values as well.

https://github.com/llvm/llvm-project/pull/111082