[PATCH] D30751: [MachineCopyForwarding] Add new pass to do register COPY forwarding at end of register allocation.

[Geoff Berry via Phabricator via llvm-commits]

gberry added a comment.

@qcolombet This change is useful primarily to increase scheduling flexibility and reducing the critical path, though it does also make some COPYs unnecessary leading to their removal.  Handling just the former in the scheduler is a possibility, but it would have the drawback of not provide a benefit to OoO cores that don't do post-RA scheduling.

As to your question of why we end up with this pattern, I looked at some cases where we end up removing COPYs and saw two main causes for this:

1. virtual registers are not getting coalesced before/during RA because there is a mismatch in register-class (e.g. the source reg class is a subset of the dest reg class).
2. only a partial segment of a complex live range has a COPY that becomes dead.

In terms of overall compiler complexity, I suspect that this pass could be extended a bit (to handle subreg COPYs for example) and make the current MachineCopyPropagation unnecessary.  There may be problems doing this related to phase ordering or different optimization level pass pipelines though, I haven't investigated it thoroughly.

I will also note FWIW that changes doing something similar to this have come up at least twice before: https://reviews.llvm.org/D21455 and https://reviews.llvm.org/D20531


https://reviews.llvm.org/D30751