[clang] [clang] Emit @llvm.assume when we know the streaming mode of the function (PR #121917)

[Nicholas Guy via cfe-commits]

NickGuy-Arm wrote:

> I really don't see why we'd want to implement the optimization this way; can't we just add an instcombine for calls to `llvm.aarch64.sme.in.streaming.mode`?

If what this patch initially did was our actual goal, then that would be an option. However what we want to achieve with this is to make the state of the streaming mode known to potentially inlineable called functions, allowing for further inlining and identification/removal of dead code. Ive fixed the implementation to actually achieve this, and I'll update the description with a bit more context

https://github.com/llvm/llvm-project/pull/121917