[clang] [llvm] [BPF] Add load-acquire and store-release instructions under -mcpu=v5 (PR #108636)

[via llvm-commits]

https://github.com/eddyz87 commented:

I think this looks fine, but syntax seems like it adds to zoo we already have.
E.g. there are already instructions like `lock *(u32 *)(r1 + 0x0) &= w2` and we treat these as having `monotonic` / `memory_order_relaxed` semantics (see #107343).
It seems, to me at-least, that it is more logical to build on top of already existing syntax, e.g.:

    lock *(u64 *)(r1 + 0x0) = r2 release
    lock r2 = *(u64 *)(r1 + 0x0) acquire

Also a question regarding 3 commits in one pull request.
As far as I understand [current](https://llvm.org/docs/GitHub.html) policy, the idea is to have a separate pull request for each commit (and fork branches from one another to build a stack). Otherwise the commits are squashed. What's the idea with current pull request?

https://github.com/llvm/llvm-project/pull/108636