[PATCH] D73411: [InstCombine] Process newly inserted instructions in the correct order

[Nikita Popov via Phabricator via llvm-commits]

nikic added a comment.

@maxim-kuvyrkov: Thanks for the report. Having the preprocessed source would be great. If you have an idea which part of the generated assembly could be related to the slowdown, that would of course be great to know as well.

Regressions from this change are not entirely unexpected, as it may expose missing folds that previously worked "by accident".


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D73411/new/

https://reviews.llvm.org/D73411