[PATCH] D147732: [AMDGPU] Add type mangling for {read, write, readfirst, perm}lane intrinsics

[Yaxun Liu via Phabricator via cfe-commits]

yaxunl added a comment.

Please make sure we can consume old IR using unmangled llvm.amdgcn.permlanex16 by fixing auto upgrade. Otherwise, it may cause regressions for device libs or existing apps. Thanks.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D147732/new/

https://reviews.llvm.org/D147732