[PATCH] D44785: Lowering x86 adds/addus/subs/subus intrinsics (llvm part)

[Tomasz Krupa via Phabricator via llvm-commits]

tkrupa added a comment.

In https://reviews.llvm.org/D44785#1045917, @RKSimon wrote:

> Please can you rebase to latest? We've already introduced SplitOpsAndApply to do what you're doing with LowerVectorOpToLegalSize


I rebased it locally and it seems I'll need to partly rework it - in the meantime somebody added a pattern for matching signed saturation and I'll need to adjust to it.


https://reviews.llvm.org/D44785