Apple | Software Engineering Intern, GPU Compilers | Coding | Onsite

This was a combination of coding and LLVM IR optimization. The coding one was LRU Cache and the llvm question is written below. LLVM IR question Optimize as much as possible: define i32 @kernel(i32 %p, i32 %n, i1 %flag) { entry: %i = alloca i32 store i32 0, i32 %i br label %loop loop: %iv = load i32, i32 %i %cmp = icmp slt i32 %iv, %n br i1 %cmp, label %body, label %exit body: %a = load i32, i32 %p %b = load i32, i32...