[all-commits] [llvm/llvm-project] 2b529a: [mlir] Removed TanHOp lowering from ConvertStandar...

[Marcel Koester via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 2b529a396d73e793615ce2cca525d68f9f738566
      https://github.com/llvm/llvm-project/commit/2b529a396d73e793615ce2cca525d68f9f738566
  Author: Marcel Koester <marcel.koester at dfki.de>
  Date:   2020-03-25 (Wed, 25 Mar 2020)

  Changed paths:
    M mlir/lib/Conversion/GPUCommon/OpToFuncCallLowering.h
    M mlir/lib/Conversion/GPUToNVVM/LowerGpuOpsToNVVMOps.cpp
    M mlir/lib/Conversion/GPUToROCDL/LowerGpuOpsToROCDLOps.cpp
    M mlir/lib/Conversion/StandardToLLVM/StandardToLLVM.cpp
    M mlir/test/Conversion/StandardToLLVM/convert-to-llvmir.mlir

  Log Message:
  -----------
  [mlir] Removed TanHOp lowering from ConvertStandardToLLVM since there is no reasonable TanH representation in LLVM.

Summary: The current ConvertStandardToLLVM phase lowers the standard TanHOp to function calls to external tanh symbols. However, this leads to misunderstandings since these external symbols are not defined anywhere. This commit removes the TanHOp lowering functionality from ConvertStandardToLLVM, adapts the LowerGpuOpsToNVVMOps and LowerGpuOpsToROCDLOps passes and adjusts the affected test cases.

Reviewers: mravishankar, herhut

Subscribers: jholewinski, mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, csigg, arpith-jacob, mgester, lucyrfox, aartbik, liufengdb, Joonsoo, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D75509