[OpenMP-OPT] Remove limit for heap to stack conversions of __kmpc_alloc_shared allocations

[doru1004]

Summary

This patch removes the limit on heap to stack conversions for allocations coming from __kmpc_alloc_shared allocations.

[jdoerfert]

Make a test for the attributor/openmp-opt, also don't use O2 in tests, the IR only test is sufficient.

[doru1004]

In D148849#4285236, @jdoerfert wrote:

Make a test for the attributor/openmp-opt, also don't use O2 in tests, the IR only test is sufficient.

I removed the clang test since it wasn't testing anything new.

[jhuber6]

LG

[doru1004]

Commit: 1a58c3d601b4c982afeb714c3a6c4be4d787cbf1

llvm/lib/Transforms/IPO/AttributorAttributes.cpp

if (Value *Align = getAllocAlignment(AI.CB, TLI)) {
                           << "\n");
         AI.Status = AllocationInfo::INVALID;
         Changed = ChangeStatus::CHANGED;
         continue;
       }
     }
 
     std::optional<APInt> Size = getSize(A, *this, AI);
-    if (MaxHeapToStackSize != -1) {
+    if (AI.LibraryFunctionId != LibFunc___kmpc_alloc_shared &&
+        MaxHeapToStackSize != -1) {
       if (!Size || Size->ugt(MaxHeapToStackSize)) {
         LLVM_DEBUG({
           if (!Size)
             dbgs() << "[H2S] Unknown allocation size: " << *AI.CB << "\n";
           else
             dbgs() << "[H2S] Allocation size too large: " << *AI.CB << " vs. "
                    << MaxHeapToStackSize << "\n";
         });

llvm/test/Transforms/Attributor/heap_to_stack_gpu.ll

bb:
   %i = tail call noalias ptr @__kmpc_alloc_shared(i64 4)
   store ptr %i, ptr @Gtl, align 8
   %i1 = load ptr, ptr @Gtl, align 8
   call void @usei8(ptr nocapture nofree %i1) nocallback nosync nounwind willreturn
   tail call void @__kmpc_free_shared(ptr %i, i64 4)
   ret void
 }
 
+define void @convert_large_kmpc_alloc_shared() {
+; CHECK-LABEL: define {{[^@]+}}@convert_large_kmpc_alloc_shared() {
+; CHECK-NEXT:  bb:
+; CHECK-NEXT:    [[I_H2S:%.*]] = alloca i8, i64 256, align 1, addrspace(5)
+; CHECK-NEXT:    [[MALLOC_CAST:%.*]] = addrspacecast ptr addrspace(5) [[I_H2S]] to ptr
+; CHECK-NEXT:    tail call void @usei8(ptr noalias nocapture nofree [[MALLOC_CAST]]) #[[ATTR7]]
+; CHECK-NEXT:    ret void
+;
+bb:
+  %i = tail call noalias ptr @__kmpc_alloc_shared(i64 256)
+  tail call void @usei8(ptr nocapture nofree %i) nosync nounwind willreturn
+  tail call void @__kmpc_free_shared(ptr %i, i64 256)
+  ret void
+}
+
 
 ;.
 ; CHECK: attributes #[[ATTR0:[0-9]+]] = { nounwind willreturn }
 ; CHECK: attributes #[[ATTR1:[0-9]+]] = { nofree nosync willreturn }
 ; CHECK: attributes #[[ATTR2:[0-9]+]] = { nofree nounwind }
 ; CHECK: attributes #[[ATTR3]] = { noreturn }
 ; CHECK: attributes #[[ATTR4:[0-9]+]] = { nocallback nofree nosync nounwind willreturn memory(argmem: readwrite) }
 ; CHECK: attributes #[[ATTR5]] = { norecurse }
 ; CHECK: attributes #[[ATTR6]] = { nounwind }
 ; CHECK: attributes #[[ATTR7]] = { nosync nounwind willreturn }
 ; CHECK: attributes #[[ATTR8]] = { nocallback nosync nounwind willreturn }
 ;.
 ;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
 ; CGSCC: {{.*}}
 ; TUNIT: {{.*}}