[PATCH] D93511: [Sink] Process basic blocks with a single successor

Matt Arsenault via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Mon Jan 4 11:10:06 PST 2021


arsenm added a comment.

In D93511#2470261 <https://reviews.llvm.org/D93511#2470261>, @MaskRay wrote:

> In D93511#2470063 <https://reviews.llvm.org/D93511#2470063>, @arsenm wrote:
>
>> Is there a reason to actually do this sinking? Normally the blocks would be merged anyway
>
> As the llc tests show the blocks cannot always be merged. I think the main motivation is not "do sinking in this case", but rather "delete an unneeded condition".

This doesn't show that? This just shows testing artifacts. It's still legal to do the sinking. I'm just saying I don't see why this is advantageous over just letting the sink pass run with the expectation that simplifycfg cleaned up trivial branches earlier


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D93511/new/

https://reviews.llvm.org/D93511



More information about the llvm-commits mailing list