[PATCH] D54649: [FPEnv] Rough out constrained FCmp intrinsics

Mon Jan 14 07:08:39 PST 2019

cameron.mcinally added a comment.

Bah, the following patch poked holes in this Diff:

  ------------------------------------------------------------------------
  r350421 | spatel | 2019-01-04 11:48:13 -0600 (Fri, 04 Jan 2019) | 18 lines

  [x86] lower extracted fadd/fsub to horizontal vector math; 2nd try

  The 1st try for this was at rL350369, but it caused IR-level diffs because
  our cost models differentiate custom vs. legal/promote lowering. So that was
  reverted at rL350373. The cost models were fixed independently at rL350403,
  so this is effectively the same patch as last time.

  Original commit message:
  This would show up if we fix horizontal reductions to narrow as they go along,
  but it's an improvement for size and/or Jaguar (fast-hops) independent of that.

  We need to do this late to not interfere with other pattern matching of larger
  horizontal sequences.

  We can extend this to integer ops in a follow-up patch.

  Differential Revision: https://reviews.llvm.org/D56011

r350421 is now custom lowering FADD and FSUB, and it breaks the generic Custom lowering changes I made to handle STRICT nodes.

Perhaps we should wait until D55506 <https://reviews.llvm.org/D55506> has landed, since that would allow us to handle STRICT_FSETCC directly in the backend? I'll dig into this more and post soon...

Repository:
  rL LLVM

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D54649/new/

https://reviews.llvm.org/D54649