[clang] [llvm] [SPIRV][HLSL] Add lowering of `rsqrt` to SPIRV (PR #95849)

[Farzon Lotfi via cfe-commits]

================
@@ -0,0 +1,29 @@
+; RUN: llc -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 %}
+
+; CHECK: OpExtInstImport "GLSL.std.450"
+
+define noundef float @rsqrt_float(float noundef %a) {
+entry:
+; CHECK: %[[#]] = OpExtInst %[[#]] %[[#]] InverseSqrt %[[#]]
----------------
farzonl wrote:

you aren't doing any type checks so I can't confirm if this is f32 or not.

after line 4 you should have something like:
```
; CHECK-DAG: %[[#float_32:]] = OpTypeFloat 32
; CHECK-DAG: %[[#float_16:]] = OpTypeFloat 16
; CHECK-DAG: %[[#float_16:]] = OpTypeFloat 64
```
If you had vec4 types they would look like

```
; CHECK-DAG: %[[#vec4_float_32:]] = OpTypeVector %[[#float_32]] 4
; CHECK-DAG: %[[#vec4_float_16:]] = OpTypeVector %[[#float_16]] 4
; CHECK-DAG: %[[#vec4_float_64:]] = OpTypeVector %[[#float_64]] 4
```

Then on line 8 it could look somethin like this:
```
; CHECK: %[[#arg0:]] = OpFunctionParameter %[[#]]
; CHECK: %[[#]] = OpExtInst %[[#float_32]] %[[#op_ext_glsl]] InverseSqrt %[[#arg0]]
```

https://github.com/llvm/llvm-project/pull/95849