[llvm] [AMDGPU] Fix a potential wrong return value indicating whether a pass modifies a function (PR #88197)

[Shilei Tian via llvm-commits]

shiltian wrote:

> We should also break out of the above loop on the first non-alloca inst

I'll do it in a follow-up patch.

https://github.com/llvm/llvm-project/pull/88197