[llvm] [NVPTX] Improve device function byval parameter lowering (PR #129188)

[Alex MacLean via llvm-commits]

================
@@ -411,15 +411,9 @@ static void adjustByValArgAlignment(Argument *Arg, Value *ArgInParamAS,
     for (User *CurUser : Ctx.InitialVal->users()) {
       if (auto *I = dyn_cast<LoadInst>(CurUser)) {
         Loads.push_back({I, Ctx.Offset});
-        continue;
-      }
-
-      if (auto *I = dyn_cast<BitCastInst>(CurUser)) {
-        Worklist.push({I, Ctx.Offset});
-        continue;
-      }
-
-      if (auto *I = dyn_cast<GetElementPtrInst>(CurUser)) {
+      } else if (isa<BitCastInst>(CurUser) || isa<AddrSpaceCastInst>(CurUser)) {
+        Worklist.push({cast<Instruction>(CurUser), Ctx.Offset});
+      } else if (auto *I = dyn_cast<GetElementPtrInst>(CurUser)) {
----------------
AlexMaclean wrote:

Fixed!

https://github.com/llvm/llvm-project/pull/129188