[llvm] [KernelInfo] Implement new LLVM IR pass for GPU code analysis (PR #102944)

[Joel E. Denny via llvm-commits]

================
@@ -824,6 +825,11 @@ void AMDGPUTargetMachine::registerPassBuilderCallbacks(PassBuilder &PB) {
           PM.addPass(AMDGPULowerModuleLDSPass(*this));
         if (EnableAMDGPUAttributor && Level != OptimizationLevel::O0)
           PM.addPass(AMDGPUAttributorPass(*this));
+        if (!NoKernelInfoEndLTO) {
----------------
jdenny-ornl wrote:

See [this comment](https://github.com/llvm/llvm-project/pull/102944#discussion_r1797494663) and [these docs](https://github.com/llvm/llvm-project/blob/d3beccfe9eb34636cd0015b867d6fcda8fa6ea26/llvm/docs/KernelInfo.rst?plain=1#L43-L63).

https://github.com/llvm/llvm-project/pull/102944