[PATCH] D116970: a better profi rebalancer

[Sergey Pupyrev via Phabricator via llvm-commits]

spupyrev added a comment.

In D116970#3236061 <https://reviews.llvm.org/D116970#3236061>, @wenlei wrote:

>> I see up to 0.5% perf on some (large) binaries, e.g., clang-10 and gcc-8.
>
> Great results! And the change also makes sense. Was this with AutoFDO or CSSPGO or InstrPGO? How does overall count quality look with this change?
>
> We can also give it a try on server workloads.

The numbers are for CSSPGO. We don't have dangling/unknown blocks in AutoFDO so it is irrelevant there. For InstrPGO, we don't run profi, if I am not mistaken?

As discussed offline, let's wait for the second (related) diff to be ready before real-world testing. This diff alone might not be enough to see a difference.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D116970/new/

https://reviews.llvm.org/D116970