[llvm] [AMDGPU] PromoteAlloca: reject known out-of-bounds index (PR #139700)

[Robert Imschweiler via llvm-commits]

================
@@ -32,7 +32,7 @@ define amdgpu_kernel void @test_private_array_ptr_calc(ptr addrspace(1) noalias
   %a = load i32, ptr addrspace(1) %a_ptr, !range !0, !noundef !{}
   %b = load i32, ptr addrspace(1) %b_ptr, !range !0, !noundef !{}
   %result = add i32 %a, %b
-  %alloca_ptr = getelementptr inbounds [16 x i32], ptr addrspace(5) %alloca, i32 0, i32 %b
+  %alloca_ptr = getelementptr inbounds [16 x i32], ptr addrspace(5) %alloca, i32 1, i32 %b
----------------
ro-i wrote:

I changed this before and did reset it in my last commit. So in comparison to main, it's now unchanged (see https://github.com/llvm/llvm-project/pull/139700/files)

https://github.com/llvm/llvm-project/pull/139700