[llvm] [AMDGPU][LowerModuleLDS] Handle non-dynlds zero sized type (PR #84742)

[Matt Arsenault via llvm-commits]

================
@@ -1398,8 +1398,8 @@ class AMDGPULowerModuleLDS {
           LDSVarsToTransform.begin(), LDSVarsToTransform.end()));
 
       for (GlobalVariable *GV : Sorted) {
-        OptimizedStructLayoutField F(GV,
-                                     DL.getTypeAllocSize(GV->getValueType()),
+        unsigned Size = DL.getTypeAllocSize(GV->getValueType());
+        OptimizedStructLayoutField F(GV, Size ? Size : 1,
----------------
arsenm wrote:

I think the correct solution is to not add it to the layout, and just track there was this unknown size object. All the 0 sized objects are supposed to receive the same address (which we would want at the end of the known allocation)

https://github.com/llvm/llvm-project/pull/84742