[PATCH] D106056: [CVP] processSwitch: Remove default case when switch cover all possible values.

Mon Sep 27 05:44:45 PDT 2021

junparser added a comment.

In D106056#3024102 <https://reviews.llvm.org/D106056#3024102>, @alexfh wrote:

> In D106056#3024018 <https://reviews.llvm.org/D106056#3024018>, @junparser wrote:
>
>> In D106056#3023813 <https://reviews.llvm.org/D106056#3023813>, @junparser wrote:
>>
>>> In D106056#3023777 <https://reviews.llvm.org/D106056#3023777>, @xbolva00 wrote:
>>>
>>>> In D106056#3023709 <https://reviews.llvm.org/D106056#3023709>, @alexfh wrote:
>>>>
>>>>> We'd appreciate a prompt fix or a workaround. If neither is possible, please revert the commit while working on a proper solution.
>>>>
>>>> Well, it is not "common" code, switch 1000 cases. But easy fix is probably to introduce some reasonable limit when to bail out.
>>>
>>> switch 1000 cases is not the issue, use loop induction variable as condition is.  It cause LVI recursive call SolveBlockValue.
>>
>> @xbolva00  I was wrong.  The compiler time regression is caused by  CodeGen Passes rather than CVP or LVI.
>
> This conclusion corresponds with the profile in my earlier comment (e.g. this line: `80.96% (anonymous namespace)::MachineBlockPlacement::buildLoopChains`).

However,  i did not hit the regression with aarch64 target.  I'll revert this first.

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D106056/new/

https://reviews.llvm.org/D106056