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

[Matt Arsenault via llvm-commits]

================
@@ -1410,3 +1410,18 @@ unsigned GCNTTIImpl::getPrefetchDistance() const {
 bool GCNTTIImpl::shouldPrefetchAddressSpace(unsigned AS) const {
   return AMDGPU::isFlatGlobalAddrSpace(AS);
 }
+
+void GCNTTIImpl::collectKernelLaunchBounds(
+    const Function &F,
+    SmallVectorImpl<std::pair<StringRef, int64_t>> &LB) const {
+  auto MaxNumWorkgroups = ST->getMaxNumWorkGroups(F);
----------------
arsenm wrote:

no auto 

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